Their support told me that soon it will be possible to track guitar Amps (I believe in a next update of their Amp Plugin) which I find very intriguing.. but now I watched the video until the end and i have to say thank you for saving my time )
You can buy the STL Plugins outright including the expansions. just a "hidden" option at the bottom of the pricing section. That being said, I think this tool is for people like myself that DON'T already have a ton of analog gear and want to use signal chains from people who are willing to provide them. For instance, I didn't even buy the Tracer potion of this tool for that reason since I know I'll be using presets over creating them.
The gem of this plugin is with the presets and the controls of the preamp unit. This video focuses only on Tracer which is a shame I actually have never used tracer but absolutely love this plugin. If you have the analog gear then use it, if not, this will get you closer than any of your channel strip emus. Also isn’t this the same sampling tech as acustica?
...that you likely wouldnt be able to use without your electricity subscription 😂..UA-cam comment also made possible by your internet service subscription
One more note. With the Color module ON, there is a 10hz high cut and 24.8hz bump at 1.4dB. Therefore, it may be why you heard the low frequency bleend, in the null test. You might try disabling the Color module.
I don’t do subscriptions but I think the Plugin Alliance is pretty good. After one year you get to keep 10. Seems a reasonable deal to me. But I don’t use that many plugins, the once’s I have are more than enough.
Would be interesting to see how control hub stacks against Neural Amp Modeller (the free guitar rig capturing software) some people have captured their outboard EQs on there. I don’t think it’s able to capture the response of dynamic processors though….
Id suspect NAM when trained at 1000 epochs can't be bested by anything but the real hardware. I can assure you that its null tests even at 100 epoch wont leave behind the entire bass and treble range.
Okay, I know that it's not "Sandstorm," but what a great sounding track, White Sea! This technology will probably improve over time. Luckily, I own all my plug-ins...
@@RicoLee27 Better put is I listen to UA-cam in the car. And if there's something I need to see, I go back and look later. And that's pretty rare. Especially with shows about audio engineering. I'm the guy that moves the plugin so the top half is off the screen so I don't cheat and use my eyes instead of my ears.
Even though it's definitely not 100% there, it's close enough to use in mixes for some stuff and repeatability and recall. I personally use my traces more for a hint of analog vibe and saturation which I made from my own console and hardware, it's a cool tool. It's not like any SSL EQ is a 100% accurate representation of a real console, that's how I consider this plugin as well.
to comapre with the inverted signal you should correct the phase between both. waves inphase plugin can do this easy. you save waveform on osciloscope view in plugin can also test in demo . maybe the stl do general phase shifts
During the A-B test, I was actually tricked into thinking that the STL was the analog gear because in my room on my Kali's it sounding better LOL and then I saw the track names and thought, "out shit, I was wrong" lol. They definitely are different. The test doesn't prove STL is better, it just means that in that moment with that particular track, I liked what it did more. I'm sure in other instances the analog gear sounds better. It just depends on the context and the material you're pushing through the gear.
i spent hours tracing both plugin chains as well hardware chains, and while it can come close some of the time, often times it is way off. When it does come close, its still missing something magic. You can hear it in the examples you provided, Sandstorm, by Darude. The analog "depth" and color is just not there. It sounds good, but yah, its not it. However... i do think in another few years we will be there. I have faith in the dev's, they make great products and i think the tech can evolve greatly. Looking forward to it
Exactly. The ControlHub clearly flattens out the sound - it lacks dimension. I could hear a difference even on Macbook Pro laptop speakers - the real analogue equipment just does sound better.
I think it got it's own eq algorithm and the analyze function just tries to match it's eq to what was captured. It won't be the same or won't null but it might come close. Just based on the observations in the video. And you are right.... what is the point if an impulse response does it better.
I would love to hear if you could play with the STL EQ during the null test to see whether there is some kind of constant offset that could make it closer.
I've tried, you can come closer for sure with a little bit of fader movement and high/low shelves, but it will pretty much never really null out. Overall, it feels darker then the original hardware.
I think there is a big difference between an impulse response that is an actual sample of the chain, and analyzing a sound and recreating that with internal processing. The latter is a lot harder to do. It's like the difference between sampling a room for the reverb and trying to recreate that same sound using an algorythmic reverb. Unless I misunderstood how the STL Controlhub works, but that's what I got from it. It analyzes the sound, determines which processors are needed and then tries to get as close as possible with the settings. But no processor, hadware or software, is without limitations. A certain curve may not be possible.
Yes, however, I’m thinking of a test between running it through the gear when it has just booted up, and running it through the gear after it has been on for a while…
Tracer works like a Kemper Profiler that captures guitar and bass amps. My Kemper can capture a guitar amp from clean to scream. BTW, STL sells Kemper packs for guitar and bass players with or without effects (compressors, overdrives, reverbs and modulation)
Kemper doesn't do it well enough to use for EQ or any studio grade effect for that matter. It doesnt process the top or bottom end fully, and some of the newest stuff out there, like Tonex do it way better. Honestly what they are doing is essentially the same as Tonex, and the same for STL here. It's all like Kemper, just way more advanced. The other issue with Kemper is it heavily compresses the dynamics, and everything feels really brickwalled. I was really hoping that Kemper would up the game by putting out and updated algorithm that didnt compress or chop off the frequencies, but instead they made the knobs work more like the real amps. I havent tried the update, but I no longer have my hands on a Kemper.
I’ve been waiting for this review by you. Great work. I’m definitely testing out this plug-in. It’s not perfect but definitely gets a close result. I wish you tried other genres to hear the different dynamics.
listening in a tuned room with a sub all set to Dolby Atmos mixing standards and yes, the lows are less warm in the STL. The analog has more depth and warmth in the low frequencies. It's close, but not exact.
In my experience it doesn't really capture your hardware. For example: if you try to capture two different VCA comps that sound very different, the two different "captures" end up sounding exactly the same (i.e identical transfer curve and envelope). It's like the plugin has a bunch of pre-programmed models, and it chooses the one that sounds the closest to the hardware you are trying to capture. Don't know if it has improved since it was first released though.
The magic 🪄 of this plugin is in the color module. Just raise the preamp knob and the drive knob a little bit. The drive knob a little bit more. - money 💵
Yup, I think that is where it really shines. Colors your get from tracing different pre amps will get you 50 percent there. It put a smile on my face. On the compressors that I traced, well it is .. almost there.too. The positive side is that it's light on my CPU. Tracing "heavy on CPU" plug ins is also a big plus.
This is the same technology used on the "Kemper Profiling Amp" ! For marketing sake(sales) They're overstating what the plugin can do.. in the right hands it might get you close.. the one thing I noticed on this post was the plugin changed the stereo image of the full mixed audio example !
Maybe try to see the phase response of the two chains. The low end culd offset by few degrees that will not sound dramaticly different. But it will trigger in a null test. Some live guys you know might have smaart or something similar🤔 11:54
the STL track def. sounds less compressed, I mean the program said it was recalculating the dynamics. maybe because of the headroom message of -5.something :) you could increase gain by some amount and test again - but in the end the IR is a simple and well proven technique, good job on that mate :)
Every time I do a null test and things don't null, the audio I hear is pretty audible. It seems if something is just a fraction of a hair of a chinny chin chin off, you're going to get something fairly audible. Maybe I'm wrong. The question I want to know is does it still sound great? I certainly can't hear the difference on my phone, of course. :/ But that's a topic for another day... Mixer Jeff Juliano is blown away by this plug-in. Is he just trying to sell his presets? Perhaps. I put this software in the category of being "close enough", just like most software nowadays. The benefit of this plugin is entire signal chains in one plugin curated by top mix engineers, and all the advantages of digital, none of the disadvantages of analog. Think about how much time and money we all could have saved if we just used this plugin from the start of our careers? There is a real freedom in letting close enough be good enough. Everybody wants that extra 5% magic that analog gives us, but we would much rather have platinum records to our name that are CLOSE ENOUGH.
The biggest issue in my oponion is that once you start turning knobs and tweaking settings it's no longer accurate. So you have to use the presets exactly as they are to get the full effect.
This isn’t a negative. You’d want tweakability. I don’t want a snapshot of a chain of gear just to slap that exact sound of every mix, I want to be able to tweak it based on each different mix
I think you completely missed my point haha!! Yes, of course you want tweakablility, that goes without saying. My point is that once you start turning the knobs tho it's no longer a snapshot of the gear. For example, if you are using a snapshot of a GBus compressor as soon as you change any settings it's no longer a snapshot of that compressor. You have to use the snapshots as they are without adjusting anything for the snapshot to be accurate. It's the same way with ToneX. ToneX is only an accurate model of the gear if you don't adjust any settings. Once you do it's not accurate to the kit being modeled. Does that make better sense?@@faudio5566
@@faudio5566 Of course you'd want tweakability. But then why pay for software that uses presets (and preset expansions by well known producers) as its backbone?
@@faudio5566 I think you misunderstood becasue with this, that's exactly what you're getting! What I mean is, let's say you are using a snapshot of a G-bus comp. That capture was taken with certain settings. So when you load the snapshot you are loading the hardware dialed into those settings. However, once you start tweaking things you are NO LONGER using a G Bus clone! The setting you are modifying are not what they would sound like on that same compressor used in the snapshot if you adjusted the settings. The full settings weren't captured. Just the settings that were set when they captured the snapshot. Hopefully that makes sense to you. It's a difficult concept to comprehend but once you do you will see what I mean. Basically once you adjust any of the settings the sound is no longer accurate to how the piece of gear would respond with the same settings.
I also did extensive testing on this "tracer" and was going to produce a video but I found out in my test that the plug does not do what it clamis, meaning is does not sound like the hardware, not even close so I decided not to do the video.
You can use analyzers but the truth its what sounds best to you. Sometimes for instance like my guitar cab captures in a mix versus the actual mic'd cab. These subtle differences are very subtle in a full mix. I do love the sound analog. If you can afford it... do it =)
One more comment: do you ever wonder if digital had been created before analog, whether we all would be preferring the sound of digital now? Maybe all we're doing is trying to chase nostalgia, eh? Maybe we would all be longing for the good old days of clean digital EQ that doesn't smear transients. I mean, maybe the true SNAKE OIL is our own flawed perception and humanity!!
I guess the use case is professional studios and engineers use this to create pre-sets to sell to people who have the software but wouldn't be able to afford a full outboard kit. Pretty much like guitar amp profiles.
Frequency dependent phase shifts happen with analog and IIR based digital filters. I wonder if phase shifts account for some of the differences in the null test? This seems like they are trying to approximate a set of transfer functions at different amplitudes and then likely interpolate the missing info. I think this is similar to what Kempers do but would have to read up…. I don’t think these types of solutions can approximate all types of distortion. However, all of what I’m saying is pure speculation:)
IMHO - I am not a big fan of Control hub and most STL products...I find that control hub and the preset packs that they sell with it feel a little bit gimmicky. The things Ive felt with it is that the presets are not retroactive - meaning that I can't A) see what parameters have been manipulated and how much as the plug in doesn't mirror the preset and B) I am not able to see what curve I am given with said preset. Its counterintuitive and found myself wasting a lot of time on the plug in 😕
Is it just me, or have STL just shown how simple it is to do the analysis on analog gear to make a plugin? Like all these companies saying about all the rigorous testing to make the “best version of an 1176 you’ve ever heard” and it’s all rubbish and takes 5 minutes to get something that sounds kiiiinnnddaaa close, but not quite there?
I think the null test is not fair, since there could simply be phase differences. Maybe the frequency response is close to identical but the phases do not align with the original.
On your point about plugin developers reaching the limits of current technology - I think that might be true for some things, and I agree the plugin industry is largely complacent and milking people dry, but there's still a ton of opportunity for innovation and using technology more creatively IMO. I think one of the bigger problems is that the market is WAY oversaturated - it sometimes feels like there are more plugin developers than people using plugins, and very few are innovating in any meaningful way. Instead, they push out the same product 1000 other companies have already done and rely on marketing bullshit, a signature gimmick, and service models to sell their products. I'd like to see more developers focus on actually inventing entirely new things. Even when not trying to replicate analog sounds, almost every plugins is based on the basic ideas and functions of analog gear, and analog gear was designed around significant constraints that don't exist or apply for digital products.
Its very close indeed and considering not every engineer can afford the luxury of expensive outboard hardware, I think this is an excellent option not forgetting every trace can be tweaked.
Straight away the guitar sounds warmer in the anolgue version. How do you use this? I mean other than getting you back to the initial setting tweaking it from therein isn't going to be the same as tweaking it the hardware. 🤔
if youre a producer and outboard analog gear is not quite in your budget purview just yet, then stl controlhub sounds great and it gets you about as close as youre gonna get to real hardware considering your current circumstances. besides, the $15 monthly subscription fee that stl controlhub charges wasnt gonna get you any closer to finally purchasing that 40k ssl desk, lol.
I have some takes on the whole A.I. audio thing. - For some very specific applications s.a. audio restauration, de-noising, and certain types of vocal processing, (generative) A.I. tools already profoundly surpass what we can do with our "traditional" analog & digital means. We're seeing A.I. tools that process audio in completely new ways and will change the way we do things completely. I think some A.I. tools will be fully adopted by everyone and anyone in 3-5 years' time. - The A.I. stuff that mimicks existing products or wants to emulate analogue gear will develop a lot slower. Our current, "traditionally" developed digital alternatives are already quite good. It will be to impossible to enter the market with a product that does the more or less the same thing for more money. The A.I. alternatives will be quite a bit more expensive because of the R&D costs. Your new big A.I. tool will have to blow the existing stuff out of the water, and the big audio software companies are hedging their bets for now. They're not sure whether most of the market will be able and willing to pay much more for a few extra percentages. You have to remember that for most audio software companies, professional audio engineers make up only a fraction of their client base. - A.I. processes still require a lot of computational resources. Having 50-60 tracks being processed in real time with A.I. applications running on a CPU is unlikely to work. We'll probably need audio processing to move away from CPU and towards to the realm of modern graphics hardware and dedicated A.I. hardware to be able to pull that off. That's coming though! - We can expect the same struggles that we're seeing in the visual processing market: It's really hard to get an A.I. to create flawless output in any situation. For a long time to come, there will be outliers, instances where A.I. software performs badly. Even Google and Microsoft will be struggling with those issues for a few years at the least. - We'll get there eventually, that's a given. We'll probably have to wait until machine learning technology matures in other, more high-margin industries & trickles down to the pro audio space. - I'm betting that in 7 years time, there'll be no audio-technical reason to be buying analog outboard gear - but it'll be commonplace for many decades still, due to how embedded it is in some artists & engineers creative workflows. PS. this plugin seems like it seems like it does some pattern recognition and matches some settings that it thinks are close. That's not actual A.I. audio-based processing.
Agreed. Doesn't seem like AI. Just turning the knobs on its own software to get as close as it can. Like positive grid bias used to do for tone matching amps.
Perhaps not a 100% true snap shot of the hardware and trace via your null test. But. ($199 Control Hub & $99 tracer, to own a license) and the ability to trace a snap shot of real hardware and insert on to many DAW tracks - not bad of a deal. Great review.
You can honestly do the exact same thing for free, with better results I suspect with NAM (Neural Amp Modeler). I would be interested to see if there was something flawed about the tests that he was using, because I really thought this would get closer. But doing my EQs, and other effects in NAM, the null test barley has anything left over, unlike the tests here.
@@lorenmorgan1931but NAM cannot capture compression or anything with time as a variable. It does an incredible job with guitar amps though.....and probably anything that saturates.
@@jacobseal Actually, you make an extremely good point here. I wonder how hard it would be to make NAM do this, and if Steve has any plans to do so in the future.
I think you are nulling wrong. Not only does it have the same audio content, it also has to have the same db. If one the analog mix is 1db louder and invert the phase to the digital mix which is 1db quieter - it won't null. Am I right or not? (Maybe I am not, but I think I am)
One: thank God i switched to Reaper some time ago. What a relief to see and hear you using it. Two: i agree with you that this is one of the worst cases of snake oil indeed. By now, i only use plugins of reknown companies, preferably one by one and certainly not in subscription form. Keep on the great efforts and: so jealous of all your mouthwatering gear indeed😊
Would love to hear what you think of "Gold Clip". As someone who doesn't use clippers I wonder if it would interest you. I think it's the best one out there hands down (probably the most expensive too 🤣)
Listening through 13 year old LG tv speakers the STL emulation track has the snare drum popping out way too much compared to the real analog chain version.
The comparison of the "matched" chain was comprehensive but I think you missed to tweak the settings that were detected because that's a possibility you don't have with an IR.
This sort of technology is likely the wave of the future. Like the way the Kemper can capture a guitar amp behavior. But I'm not tempted enough for a subscription model.
The IR test makes me laugh so much ahahah 😂 I'd liek to see te test on Compressor...nowadays the not dynamic things are acually capturable, basically like Tonex for guitars or Quad Cortex...it's interesting thing plugin for pro audio too, but is the dynamic part the most interesting, so please make a test with compressors.
I'm very impressed that it's not aliasing... I've seen some cool results with machine learning lately.... Non-linearities.... And somehow getting really close to some cool analog stuff... But the method is totally different than impulse responses... I could see it not being accurate in the right way to know that still doing something new that's kind of exciting and different than normal algorithmic stuff... Obviously you can't use this to clone your hardware!!! But there's something going on that's cool and this plugin, I think... And there's some other machine learning stuff that's more accurate than this... at least to my ears. Like I really like some of the Tonex captures... It can really feel like you're playing a good app with all the dynamics.... Edge of breakup stuff.... Amp like. I'm wondering if a machine learning could get better for this type of application too. One thing I do like about this plugin is how tweakable it is... Like you can change all the parameters afterwards. And if it's not aliasing.... Somebody could find some use with a plugin like this. Especially if it really is capturing some of those analog like idiosyncrasies. I think it's worth investigating.
ok at 18 mins - its a impulse response poorly implemented. or its altered to jack up certain genres for bedroom producers (‘oh wow this is amazing’) using psychoacoustics?
Great video as always. Such a shame for STL that a humble IR nulled better than ControlHub! Watching this, I wondered how modeling software like IK TONEX or Acustica Nebula would fare in comparison. I also wonder whether tracing each unit separately might improve the chain. The manual says Compressor-specific tracing is good for just getting one compressor and integrating it into an existing rig. But for what it's worth, ControlHub’s compressor module only has VCA and FET topologies - not vari-mu. So, I suspect the UAD Fairchild or Kazrog True Dynamics could impersonate the Dictator better than any multi-FX options.
Yea...a lot of the plugins are the same with different skins on them when updated !! Yet I do appreciate how many of the new skins make using the plug in easier 2 use!! In some cases improvements are made from analyzing peoples use!! I think one day we may see a suite that after each recording AI will do everything for you!! When that happens I thing the human ear will only start 2 accept those created frequencies !! This would likely be the death of music as we knew it!! Mixing is an art !! Mistakes 2 one person could be a creative indifference 2 another !! A garage band or a genuine raw sound will not be understood !! It is just speculation !! It may wind up being a good thang for a while !! Then I believe people will get tired of all the automation and turn 2 acoustics again !!!
Could you please compare some of the rumoured reshelled algorithms? It would be great to expose this bs. It's the reason im considering a basic analogue mixing rig. Get some good equipment and be done with it. Rather than analysing expensive plugins that get replaced by expensive slightly "better" plugins. It's become tiresome, the more I learn.
Thank you so much for featuring my song "Rudestorm" by The Sand
😂
You’re doing gods work. That analog chain don’t lie.
Would love to see you do a video covering all the Tukan plugins for Reaper and how you think they stack up against paid plugins.
Their support told me that soon it will be possible to track guitar Amps (I believe in a next update of their Amp Plugin) which I find very intriguing.. but now I watched the video until the end and i have to say thank you for saving my time )
You can buy the STL Plugins outright including the expansions. just a "hidden" option at the bottom of the pricing section. That being said, I think this tool is for people like myself that DON'T already have a ton of analog gear and want to use signal chains from people who are willing to provide them. For instance, I didn't even buy the Tracer potion of this tool for that reason since I know I'll be using presets over creating them.
The gem of this plugin is with the presets and the controls of the preamp unit. This video focuses only on Tracer which is a shame
I actually have never used tracer but absolutely love this plugin. If you have the analog gear then use it, if not, this will get you closer than any of your channel strip emus.
Also isn’t this the same sampling tech as acustica?
My only rule in audio: if it’s subscription. Stay away
...that you likely wouldnt be able to use without your electricity subscription 😂..UA-cam comment also made possible by your internet service subscription
I know what you mean
Agreed. Look how many plugin companies are not even in business anymore.
@@phadrus Please share with us your subscription horror story..
@@danepaulstewart8464 Which also means you're no longer getting compatibility updates for your perpetual license lol
One more note. With the Color module ON, there is a 10hz high cut and 24.8hz bump at 1.4dB.
Therefore, it may be why you heard the low frequency bleend, in the null test.
You might try disabling the Color module.
So the difference in this video is because he didn’t turn that module off?
@@nagemaakte might be. I would love to see a follow up on this or if anyone else has the plugin and ran this test with it off.
A subscription is NEVER too good to be true.
@@officialWWM Stockholm syndrome
false
I don’t do subscriptions but I think the Plugin Alliance is pretty good. After one year you get to keep 10.
Seems a reasonable deal to me. But I don’t use that many plugins, the once’s I have are more than enough.
You can buy a perpetual license.
On the full chain even through YT compression It lost a lot of '3d' and what I call 'weight variance' where you can hear the bottom end 'breathe'
Big Respect for reviewing this plugin! This plugin inspire me to make a lot of hits!
Would be interesting to see how control hub stacks against Neural Amp Modeller (the free guitar rig capturing software) some people have captured their outboard EQs on there. I don’t think it’s able to capture the response of dynamic processors though….
I'd love to see that too! I love NAM and Proteus outboard captures.
Id suspect NAM when trained at 1000 epochs can't be bested by anything but the real hardware. I can assure you that its null tests even at 100 epoch wont leave behind the entire bass and treble range.
We just need them to make it stereo and figure out how to deal with it's alleged aliasing issue.@@lorenmorgan1931
Okay, I know that it's not "Sandstorm," but what a great sounding track, White Sea!
This technology will probably improve over time. Luckily, I own all my plug-ins...
8:43 “That’s the sound of my home planet.”
I almost had to pull my car over I was laughing so hard.
You watch UA-cam in the car???? Interesting generation we living in I geuss
@@RicoLee27 Better put is I listen to UA-cam in the car. And if there's something I need to see, I go back and look later. And that's pretty rare. Especially with shows about audio engineering. I'm the guy that moves the plugin so the top half is off the screen so I don't cheat and use my eyes instead of my ears.
Before you even nulled it, I knew that the bass was a big difference. I don't have the best boxes, but I know, when they start to create more boom.
Even though it's definitely not 100% there, it's close enough to use in mixes for some stuff and repeatability and recall. I personally use my traces more for a hint of analog vibe and saturation which I made from my own console and hardware, it's a cool tool. It's not like any SSL EQ is a 100% accurate representation of a real console, that's how I consider this plugin as well.
This is closer than anything I have used before including Acustica
to comapre with the inverted signal you should correct the phase between both. waves inphase plugin can do this easy. you save waveform on osciloscope view in plugin can also test in demo . maybe the stl do general phase shifts
During the A-B test, I was actually tricked into thinking that the STL was the analog gear because in my room on my Kali's it sounding better LOL and then I saw the track names and thought, "out shit, I was wrong" lol. They definitely are different. The test doesn't prove STL is better, it just means that in that moment with that particular track, I liked what it did more. I'm sure in other instances the analog gear sounds better. It just depends on the context and the material you're pushing through the gear.
i spent hours tracing both plugin chains as well hardware chains, and while it can come close some of the time, often times it is way off. When it does come close, its still missing something magic. You can hear it in the examples you provided, Sandstorm, by Darude. The analog "depth" and color is just not there. It sounds good, but yah, its not it. However... i do think in another few years we will be there. I have faith in the dev's, they make great products and i think the tech can evolve greatly. Looking forward to it
Exactly. The ControlHub clearly flattens out the sound - it lacks dimension. I could hear a difference even on Macbook Pro laptop speakers - the real analogue equipment just does sound better.
I love the way you're dissapointed by this software.. :)
man I thought the STL one sounded better without knowing which was which haha
I think it got it's own eq algorithm and the analyze function just tries to match it's eq to what was captured. It won't be the same or won't null but it might come close. Just based on the observations in the video. And you are right.... what is the point if an impulse response does it better.
I would love to hear if you could play with the STL EQ during the null test to see whether there is some kind of constant offset that could make it closer.
I've tried, you can come closer for sure with a little bit of fader movement and high/low shelves, but it will pretty much never really null out. Overall, it feels darker then the original hardware.
I think there is a big difference between an impulse response that is an actual sample of the chain, and analyzing a sound and recreating that with internal processing. The latter is a lot harder to do. It's like the difference between sampling a room for the reverb and trying to recreate that same sound using an algorythmic reverb. Unless I misunderstood how the STL Controlhub works, but that's what I got from it. It analyzes the sound, determines which processors are needed and then tries to get as close as possible with the settings. But no processor, hadware or software, is without limitations. A certain curve may not be possible.
If you ran a piece of audio through your analog gear twice, with the same settings, would it null perfectly?
Yes, however, I’m thinking of a test between running it through the gear when it has just booted up, and running it through the gear after it has been on for a while…
Tracer works like a Kemper Profiler that captures guitar and bass amps. My Kemper can capture a guitar amp from clean to scream. BTW, STL sells Kemper packs for guitar and bass players with or without effects (compressors, overdrives, reverbs and modulation)
Kemper doesn't do it well enough to use for EQ or any studio grade effect for that matter. It doesnt process the top or bottom end fully, and some of the newest stuff out there, like Tonex do it way better. Honestly what they are doing is essentially the same as Tonex, and the same for STL here. It's all like Kemper, just way more advanced. The other issue with Kemper is it heavily compresses the dynamics, and everything feels really brickwalled. I was really hoping that Kemper would up the game by putting out and updated algorithm that didnt compress or chop off the frequencies, but instead they made the knobs work more like the real amps. I havent tried the update, but I no longer have my hands on a Kemper.
I’ve been waiting for this review by you. Great work. I’m definitely testing out this plug-in. It’s not perfect but definitely gets a close result. I wish you tried other genres to hear the different dynamics.
listening in a tuned room with a sub all set to Dolby Atmos mixing standards and yes, the lows are less warm in the STL. The analog has more depth and warmth in the low frequencies. It's close, but not exact.
In my experience it doesn't really capture your hardware. For example: if you try to capture two different VCA comps that sound very different, the two different "captures" end up sounding exactly the same (i.e identical transfer curve and envelope). It's like the plugin has a bunch of pre-programmed models, and it chooses the one that sounds the closest to the hardware you are trying to capture. Don't know if it has improved since it was first released though.
Curious if it would be closer sounding if each of your components was modeled individually in the chain?
The magic 🪄 of this plugin is in the color module. Just raise the preamp knob and the drive knob a little bit.
The drive knob a little bit more.
- money 💵
Yup, I think that is where it really shines. Colors your get from tracing different pre amps will get you 50 percent there. It put a smile on my face. On the compressors that I traced, well it is .. almost there.too. The positive side is that it's light on my CPU. Tracing "heavy on CPU" plug ins is also a big plus.
This is the same technology used on the "Kemper Profiling Amp" ! For marketing sake(sales) They're overstating what the plugin can do.. in the right hands it might get you close.. the one thing I noticed on this post was the plugin changed the stereo image of the full mixed audio example !
Maybe try to see the phase response of the two chains. The low end culd offset by few degrees that will not sound dramaticly different. But it will trigger in a null test. Some live guys you know might have smaart or something similar🤔 11:54
the STL track def. sounds less compressed, I mean the program said it was recalculating the dynamics. maybe because of the headroom message of -5.something :) you could increase gain by some amount and test again - but in the end the IR is a simple and well proven technique, good job on that mate :)
Every time I do a null test and things don't null, the audio I hear is pretty audible. It seems if something is just a fraction of a hair of a chinny chin chin off, you're going to get something fairly audible. Maybe I'm wrong. The question I want to know is does it still sound great? I certainly can't hear the difference on my phone, of course. :/ But that's a topic for another day...
Mixer Jeff Juliano is blown away by this plug-in. Is he just trying to sell his presets? Perhaps. I put this software in the category of being "close enough", just like most software nowadays. The benefit of this plugin is entire signal chains in one plugin curated by top mix engineers, and all the advantages of digital, none of the disadvantages of analog.
Think about how much time and money we all could have saved if we just used this plugin from the start of our careers? There is a real freedom in letting close enough be good enough. Everybody wants that extra 5% magic that analog gives us, but we would much rather have platinum records to our name that are CLOSE ENOUGH.
The biggest issue in my oponion is that once you start turning knobs and tweaking settings it's no longer accurate. So you have to use the presets exactly as they are to get the full effect.
This isn’t a negative. You’d want tweakability. I don’t want a snapshot of a chain of gear just to slap that exact sound of every mix, I want to be able to tweak it based on each different mix
I think you completely missed my point haha!! Yes, of course you want tweakablility, that goes without saying. My point is that once you start turning the knobs tho it's no longer a snapshot of the gear. For example, if you are using a snapshot of a GBus compressor as soon as you change any settings it's no longer a snapshot of that compressor. You have to use the snapshots as they are without adjusting anything for the snapshot to be accurate. It's the same way with ToneX. ToneX is only an accurate model of the gear if you don't adjust any settings. Once you do it's not accurate to the kit being modeled. Does that make better sense?@@faudio5566
@@faudio5566 Of course you'd want tweakability. But then why pay for software that uses presets (and preset expansions by well known producers) as its backbone?
doesn't sound like we're talking about the same thing. must be some misunderstanding.@@CJ_Wolesz
@@faudio5566 I think you misunderstood becasue with this, that's exactly what you're getting! What I mean is, let's say you are using a snapshot of a G-bus comp. That capture was taken with certain settings. So when you load the snapshot you are loading the hardware dialed into those settings. However, once you start tweaking things you are NO LONGER using a G Bus clone! The setting you are modifying are not what they would sound like on that same compressor used in the snapshot if you adjusted the settings. The full settings weren't captured. Just the settings that were set when they captured the snapshot. Hopefully that makes sense to you. It's a difficult concept to comprehend but once you do you will see what I mean. Basically once you adjust any of the settings the sound is no longer accurate to how the piece of gear would respond with the same settings.
I also did extensive testing on this "tracer" and was going to produce a video but I found out in my test that the plug does not do what it clamis, meaning is does not sound like the hardware, not even close so I decided not to do the video.
You can use analyzers but the truth its what sounds best to you. Sometimes for instance like my guitar cab captures in a mix versus the actual mic'd cab. These subtle differences are very subtle in a full mix. I do love the sound analog. If you can afford it... do it =)
One more comment: do you ever wonder if digital had been created before analog, whether we all would be preferring the sound of digital now? Maybe all we're doing is trying to chase nostalgia, eh? Maybe we would all be longing for the good old days of clean digital EQ that doesn't smear transients. I mean, maybe the true SNAKE OIL is our own flawed perception and humanity!!
I guess the use case is professional studios and engineers use this to create pre-sets to sell to people who have the software but wouldn't be able to afford a full outboard kit. Pretty much like guitar amp profiles.
I would love to see if you could dail it in closer with the plugin parameters
The industry is definitely saturated with greed sentiment in context of selling something 'new'
stick to objects we can hold and touch! they will still be there when the plugin manufacturers crumble
One of my favourite sandstorm variations. Loved this so much 😊
Can someone explain to me the sandstorm references in his videos. Is it like an inside joke?
@@chadmichael_ absolutely is 😅
@@jadedragon469 how so?
@@chadmichael_ every time he uses a ref track he labels the title as darude sandstorm
@@jadedragon469 is it something that his audience picked up on overtime and started making comments about?
This was a fun video. Thanks for uploading.
Is this like ToneX for hardware mixing gear?
Those weird noises were absolutely fantastic
Frequency dependent phase shifts happen with analog and IIR based digital filters. I wonder if phase shifts account for some of the differences in the null test? This seems like they are trying to approximate a set of transfer functions at different amplitudes and then likely interpolate the missing info. I think this is similar to what Kempers do but would have to read up…. I don’t think these types of solutions can approximate all types of distortion. However, all of what I’m saying is pure speculation:)
IMHO - I am not a big fan of Control hub and most STL products...I find that control hub and the preset packs that they sell with it feel a little bit gimmicky. The things Ive felt with it is that the presets are not retroactive - meaning that I can't A) see what parameters have been manipulated and how much as the plug in doesn't mirror the preset and B) I am not able to see what curve I am given with said preset. Its counterintuitive and found myself wasting a lot of time on the plug in 😕
"there's always some snake oil in these type of things..." LOL my favorite part of the video. GREAT Video btw! 💪🏽💪🏽💪🏽🔥🔥🔥
Well, no one stops you from tweaking the buttons if it is not close enough? The thing is if you know what you want to achieve or not?
Is it just me, or have STL just shown how simple it is to do the analysis on analog gear to make a plugin? Like all these companies saying about all the rigorous testing to make the “best version of an 1176 you’ve ever heard” and it’s all rubbish and takes 5 minutes to get something that sounds kiiiinnnddaaa close, but not quite there?
I’d be interested to hear you record the piece twice through the analog chain and null those? Would it actually cancel out completely?
Thanks for the video man. Appreciated. BTW what Mic are you speaking in for this video? Looks spacey and sounds all right 😀
I think the null test is not fair, since there could simply be phase differences. Maybe the frequency response is close to identical but the phases do not align with the original.
On your point about plugin developers reaching the limits of current technology - I think that might be true for some things, and I agree the plugin industry is largely complacent and milking people dry, but there's still a ton of opportunity for innovation and using technology more creatively IMO. I think one of the bigger problems is that the market is WAY oversaturated - it sometimes feels like there are more plugin developers than people using plugins, and very few are innovating in any meaningful way. Instead, they push out the same product 1000 other companies have already done and rely on marketing bullshit, a signature gimmick, and service models to sell their products. I'd like to see more developers focus on actually inventing entirely new things. Even when not trying to replicate analog sounds, almost every plugins is based on the basic ideas and functions of analog gear, and analog gear was designed around significant constraints that don't exist or apply for digital products.
Its very close indeed and considering not every engineer can afford the luxury of expensive outboard hardware, I think this is an excellent option not forgetting every trace can be tweaked.
Then again, why not just use any modelling plugin that is also very close?
So its the Kemper of studio gear!
Should in other words, be able to capture amp profiles just like a Kemper too?? :D
Straight away the guitar sounds warmer in the anolgue version. How do you use this? I mean other than getting you back to the initial setting tweaking it from therein isn't going to be the same as tweaking it the hardware. 🤔
if youre a producer and outboard analog gear is not quite in your budget purview just yet, then stl controlhub sounds great and it gets you about as close as youre gonna get to real hardware considering your current circumstances. besides, the $15 monthly subscription fee that stl controlhub charges wasnt gonna get you any closer to finally purchasing that 40k ssl desk, lol.
Actually any penny gets you
closer to any outboard gear 😊
ALOG has swing ,, STL seems like its syncopated TOOOO accurately on time ,, but interestinggg hmmm
Great great video! And 100% consent. You are a kind of Robin Hood in the audio world!
He’s wearing the green hoodie 😂
I have some takes on the whole A.I. audio thing.
- For some very specific applications s.a. audio restauration, de-noising, and certain types of vocal processing, (generative) A.I. tools already profoundly surpass what we can do with our "traditional" analog & digital means. We're seeing A.I. tools that process audio in completely new ways and will change the way we do things completely. I think some A.I. tools will be fully adopted by everyone and anyone in 3-5 years' time.
- The A.I. stuff that mimicks existing products or wants to emulate analogue gear will develop a lot slower. Our current, "traditionally" developed digital alternatives are already quite good.
It will be to impossible to enter the market with a product that does the more or less the same thing for more money. The A.I. alternatives will be quite a bit more expensive because of the R&D costs. Your new big A.I. tool will have to blow the existing stuff out of the water, and the big audio software companies are hedging their bets for now. They're not sure whether most of the market will be able and willing to pay much more for a few extra percentages. You have to remember that for most audio software companies, professional audio engineers make up only a fraction of their client base.
- A.I. processes still require a lot of computational resources. Having 50-60 tracks being processed in real time with A.I. applications running on a CPU is unlikely to work. We'll probably need audio processing to move away from CPU and towards to the realm of modern graphics hardware and dedicated A.I. hardware to be able to pull that off. That's coming though!
- We can expect the same struggles that we're seeing in the visual processing market: It's really hard to get an A.I. to create flawless output in any situation. For a long time to come, there will be outliers, instances where A.I. software performs badly. Even Google and Microsoft will be struggling with those issues for a few years at the least.
- We'll get there eventually, that's a given. We'll probably have to wait until machine learning technology matures in other, more high-margin industries & trickles down to the pro audio space.
- I'm betting that in 7 years time, there'll be no audio-technical reason to be buying analog outboard gear - but it'll be commonplace for many decades still, due to how embedded it is in some artists & engineers creative workflows.
PS. this plugin seems like it seems like it does some pattern recognition and matches some settings that it thinks are close. That's not actual A.I. audio-based processing.
Agreed. Doesn't seem like AI. Just turning the knobs on its own software to get as close as it can. Like positive grid bias used to do for tone matching amps.
Very interesting analysis of the reason why most companies go to subscription plans.
Perhaps not a 100% true snap shot of the hardware and trace via your null test.
But. ($199 Control Hub & $99 tracer, to own a license) and the ability to trace a snap shot of real
hardware and insert on to many DAW tracks - not bad of a deal. Great review.
You can honestly do the exact same thing for free, with better results I suspect with NAM (Neural Amp Modeler). I would be interested to see if there was something flawed about the tests that he was using, because I really thought this would get closer. But doing my EQs, and other effects in NAM, the null test barley has anything left over, unlike the tests here.
@@lorenmorgan1931but NAM cannot capture compression or anything with time as a variable. It does an incredible job with guitar amps though.....and probably anything that saturates.
@@jacobseal Actually, you make an extremely good point here. I wonder how hard it would be to make NAM do this, and if Steve has any plans to do so in the future.
I think you are nulling wrong. Not only does it have the same audio content, it also has to have the same db. If one the analog mix is 1db louder and invert the phase to the digital mix which is 1db quieter - it won't null. Am I right or not? (Maybe I am not, but I think I am)
I did adjust for gain, this was the nullest I could get them to jull
@@Whiteseastudio 😲
One: thank God i switched to Reaper some time ago. What a relief to see and hear you using it. Two: i agree with you that this is one of the worst cases of snake oil indeed. By now, i only use plugins of reknown companies, preferably one by one and certainly not in subscription form. Keep on the great efforts and: so jealous of all your mouthwatering gear indeed😊
Would love to hear what you think of "Gold Clip". As someone who doesn't use clippers I wonder if it would interest you. I think it's the best one out there hands down (probably the most expensive too 🤣)
nice! darude playing like always!!
That’s the sound of my home planet ❤
Jep 😂😂😂
Listening through 13 year old LG tv speakers the STL emulation track has the snare drum popping out way too much compared to the real analog chain version.
I was in the same place. Thank you.
maybe there is a small delay of a few samples between the amalog and the stl
I did correct for latency and gain, in such a way that it made the most silent null…
do you considering make presets
Thanks for this in depth review. been waiting for someone to stretch it out and put it through paces.
The dynamics of the bass is there but the fullness and depth and perceived girth...seemed to not be there compared to the analog
I would advise everyone to try the demo and forget the analog digital bs and just listen to the sound
His home planet is calling :)
Love this man .... thank u so much
It would be interesting to see a review not on the capture technology, but on the sound of the built-in algorithms and convenience
There are alot of reviews out there because the software has been out there for ages just preset only as far as I understand.
You should try doing the exact same hardware twice in a row and compare the modeling!
The comparison of the "matched" chain was comprehensive but I think you missed to tweak the settings that were detected because that's a possibility you don't have with an IR.
You've all the analog stuff for these last 3 to 4% that tiny bit of magic. I think it comes really close, but it fails in about 3 to 4% so yeah...
This plugin significantly simplifies and speeds up my mixing. I really like it, but I'm not much of a producer.
You're sorta missing the point here, the STL sets up a chain that can be adjusted, tweaked, elements switched on and off, and so on . . .
This sort of technology is likely the wave of the future. Like the way the Kemper can capture a guitar amp behavior.
But I'm not tempted enough for a subscription model.
It adds a noticeable hum to the mix.
👍 Oil Snake? Thanks.
Well done, indeed
I can hear it's not the same but at the same time it is not too far neither...Maybe with a few tweaks you can get it closer...
The IR test makes me laugh so much ahahah 😂 I'd liek to see te test on Compressor...nowadays the not dynamic things are acually capturable, basically like Tonex for guitars or Quad Cortex...it's interesting thing plugin for pro audio too, but is the dynamic part the most interesting, so please make a test with compressors.
its absolutely possible with ai tech now, we’re waiting for proper real time plugins now. tonex and nam are already doing great job at it
AI is just another word for software. It isn't magic, it is The Force. It's just software.
@@kadiummusiclike ozone lol
@@kadiummusicyea it's essentially it's just a pattern and prediction software
@@kadiummusic explain to us how it works??
I'm very impressed that it's not aliasing...
I've seen some cool results with machine learning lately.... Non-linearities.... And somehow getting really close to some cool analog stuff... But the method is totally different than impulse responses... I could see it not being accurate in the right way to know that still doing something new that's kind of exciting and different than normal algorithmic stuff...
Obviously you can't use this to clone your hardware!!! But there's something going on that's cool and this plugin, I think... And there's some other machine learning stuff that's more accurate than this... at least to my ears. Like I really like some of the Tonex captures... It can really feel like you're playing a good app with all the dynamics.... Edge of breakup stuff.... Amp like. I'm wondering if a machine learning could get better for this type of application too.
One thing I do like about this plugin is how tweakable it is...
Like you can change all the parameters afterwards. And if it's not aliasing.... Somebody could find some use with a plugin like this. Especially if it really is capturing some of those analog like idiosyncrasies. I think it's worth investigating.
If doesn’t null, is not the same, but even , I like it
impulse responses? that was my first thought how this works. (at 4 mins in lol)
ok at 18 mins - its a impulse response poorly implemented. or its altered to jack up certain genres for bedroom producers (‘oh wow this is amazing’) using psychoacoustics?
Great video as always. Such a shame for STL that a humble IR nulled better than ControlHub! Watching this, I wondered how modeling software like IK TONEX or Acustica Nebula would fare in comparison.
I also wonder whether tracing each unit separately might improve the chain. The manual says Compressor-specific tracing is good for just getting one compressor and integrating it into an existing rig. But for what it's worth, ControlHub’s compressor module only has VCA and FET topologies - not vari-mu. So, I suspect the UAD Fairchild or Kazrog True Dynamics could impersonate the Dictator better than any multi-FX options.
Yea...a lot of the plugins are the same with different skins on them when updated !! Yet I do appreciate how many of the new skins make using the plug in easier 2 use!! In some cases improvements are made from analyzing peoples use!! I think one day we may see a suite that after each recording AI will do everything for you!! When that happens I thing the human ear will only start 2 accept those created frequencies !! This would likely be the death of music as we knew it!! Mixing is an art !! Mistakes 2 one person could be a creative indifference 2 another !! A garage band or a genuine raw sound will not be understood !! It is just speculation !! It may wind up being a good thang for a while !! Then I believe people will get tired of all the automation and turn 2 acoustics again !!!
Could you please compare some of the rumoured reshelled algorithms? It would be great to expose this bs.
It's the reason im considering a basic analogue mixing rig. Get some good equipment and be done with it. Rather than analysing expensive plugins that get replaced by expensive slightly "better" plugins. It's become tiresome, the more I learn.
I thought I was crazy i used it and it killed the vibe of my mix went back to fabfilter....I'm disappointed it's another snake oil
Wytze - another great term for ‘mierenneuken’ is ‘splitting hairs’ 😊
None of the companies are able to capture the full analog magic