Sonible smart:limit

[fairlyclose]

4. Despite having all those publishing target settings, it doesn't actually meet those targets. It makes a loud master then tells you how much the publisher will turn it down by. Not what I expected.

I think the idea here is for you to get the sound you are after first and target being louder than the publisher requirement so that they will turn it down as this does not effect the actual sound that gets streamed. But if you have the sound you want and are softer the service will turn it up and possibly limit the track themselves and stuff up the sound you want. They discuss this a bit in the manual

[bduffy]

4. Despite having all those publishing target settings, it doesn't actually meet those targets. It makes a loud master then tells you how much the publisher will turn it down by. Not what I expected.

I think the idea here is for you to get the sound you are after first and target being louder than the publisher requirement so that they will turn it down as this does not effect the actual sound that gets streamed. But if you have the sound you want and are softer the service will turn it up and possibly limit the track themselves and stuff up the sound you want. They discuss this a bit in the manual

For sure. Mastering to around -9dB LUFS and letting Spotify turn it down is a standard practice and good for covering your bases. That's fine.

But when I submit a mix for broadcast, it has to hit -24 LUFS. They don't turn it down, it gets sent back! So it seems a little odd to me that the broadcast presets didn't optimize the output, like Waves WLM Plus does. The plugin knows the number, why not just...apply it to the output?

It seems odd, if its optimizing for a delivery target, but it doesn't adjust the gain structure to meet that target.

I'm awaiting a response from sonible on this. It's just weird to me that they designed this assuming the master is always going to be turned down. Apart from that, I love this limiter.

[fairlyclose]

4. Despite having all those publishing target settings, it doesn't actually meet those targets. It makes a loud master then tells you how much the publisher will turn it down by. Not what I expected.

I think the idea here is for you to get the sound you are after first and target being louder than the publisher requirement so that they will turn it down as this does not effect the actual sound that gets streamed. But if you have the sound you want and are softer the service will turn it up and possibly limit the track themselves and stuff up the sound you want. They discuss this a bit in the manual

For sure. Mastering to around -9dB LUFS and letting Spotify turn it down is a standard practice and good for covering your bases. That's fine.

But when I submit a mix for broadcast, it has to hit -24 LUFS. They don't turn it down, it gets sent back! So it seems a little odd to me that the broadcast presets didn't optimize the output, like Waves WLM Plus does. The plugin knows the number, why not just...apply it to the output?

It seems odd, if its optimizing for a delivery target, but it doesn't adjust the gain structure to meet that target.

I'm awaiting a response from sonible on this. It's just weird to me that they designed this assuming the master is always going to be turned down. Apart from that, I love this limiter.

thanks for that explanation - perhaps a button "apply correction"
I am pretty impressed even though what I do has a much larger dynamic range than is usual in commercial musics. I am going to try the reference track system and see how that goes

Reference track works a treat so this is a buy for me. A lot of what I do is field recording or field recording based or contemporary classical so there can be huge dynamic range. This lets me maintain that effective range measured against something similar that I like. I can easily pull back on the recommendation to keep greater dynamics. Gonna be a real time saver

[DCrown]

Watched an interesting video review of sonible, elevate, Kclip and more. I also think AI could be improved and why level of mix is not automatically adjusted to target indeed?! I agree the analysing procedure is too short. It seems to be for very repetitive music only.

[pierb]

No opinions on the product itself yet (barely used it) and will have to compare it to my usual favorites (Pro L2, Elephant, Elevate and Ozone). Looks like it has some interesting ideas, but it'll be an uphill battle against the competition for sure.

So how does it compare to Elevate, L2, etc?

[fairlyclose]

Watched an interesting video review. I also think AI could be improved and why level of mix is not automatically adjusted to target indeed?! I agree the analysing procedure is too short. It seems to be for very repetitive music only.

You dont necessarily want level auto adjusted, but a button/setting for when you do might be nice
The point about learning time being too short needs to be justified against performance across a range of material. Bound to be some edge cases as with all machine learning. Knowing where they are takes experience with those cases which you might never run into. Or they might be bread and butter for you. Find out by testing with the demo

[bduffy]

I'm really hoping for a nice .5 update on this. It's so close to being the best VST limiter out there, just needs to hit those targets. And have oversampling.

[MadDogE134]

@ sl1200mk2 the openGL seems to be an nVidia problem in the latest driver. if you revert back to the last driver... 472.12 i think it is... no problem and NO the graphic problem does not affect the functionality of the plugins... you just don't get the pretty graphics. i only saw the missing graphics on the smart:reverb and the smart:comp and it is a simple fix you can do yourself. it is not wise to ALWAYS install video drivers unless you actually have a problem with your card or it's performance. staying on the 'bleeding' edge has it's share of problems.

if you know anything about video drivers... especially nVidia... new drivers usually foul something up mostly with gaming that rely heavily on openGL and overlays. i would surmise that if Sonible left well enough alone the 'next' nVidia driver would fix nVidia's own problem. do you think developers should jump through hoops every time a video driver is faulty? i used to compete online and it is a well known fact that it is not wise to update a video driver just because a new one comes out lol

cheers

p.s. i personally love the new limiter. very impressive imo

[MogwaiBoy]

But when I submit a mix for broadcast, it has to hit -24 LUFS.

Sometimes I wish all music had to be mastered to -24 LUFS... for certain kinds of music the dynamics would be insanely awesome!

[fairlyclose]

There am oversampling
posted on gearspace

Elisabeth Frauscher (sonible.com)
Dec 13, 2021, 16:50 GMT+1

Since the smart:limit plugin is a true-peak limiter it performs of course oversampling and uses an anti-aliasing filter for downsampling.

Best,
Elisabeth

[plexuss]

But when I submit a mix for broadcast, it has to hit -24 LUFS.

Sometimes I wish all music had to be mastered to -24 LUFS... for certain kinds of music the dynamics would be insanely awesome!

I've posted this before: In my experiments to find the optimal balance of loudness, ISP mitigation and dynamics, I've found -16 LUFS is a good target. Sometimes I got a bit over depending on the material. I think for good quality music audio, -16 is optimal because it will typically get a decent amount of volume while usually just normalizing amplitude, minimizing clipping/limiting. So with -16 I never get ISP warnings, get a decent amount of loudness which is better for usability (eg. volume control setting) and still preserves a good amount of peak dynamics and LRA. Here's a couple examples:


https://soundcloud.com/musicofplexus/blue-dream

[MadDogE134]

i personally like a gnat's ass under -14 LUFS i think it is... plenty of umph but not so much as to push things

[bduffy]

There am oversampling
posted on gearspace

Elisabeth Frauscher (sonible.com)
Dec 13, 2021, 16:50 GMT+1

Since the smart:limit plugin is a true-peak limiter it performs of course oversampling and uses an anti-aliasing filter for downsampling.

Best,
Elisabeth

Ah! I searched variations of "oversampling" in the manual and it didn't appear to be there. Good to know.

[sl1200mk2]

@ sl1200mk2 the openGL seems to be an nVidia problem in the latest driver. if you revert back to the last driver... 472.12 i think it is... no problem and NO the graphic problem does not affect the functionality of the plugins... you just don't get the pretty graphics.

Thanks. I stated that the plugins were still 'functional' in that the controls still worked as you would expect. That said, it's my opinion that the entire frequency spectrum missing on Smart:EQ is a fairly substantial issue not to be fixed for a couple of months and counting given it's a largely GUI based EQ and not just a few knobs and/or sliders. YMMV.

i would surmise that if Sonible left well enough alone the 'next' nVidia driver would fix nVidia's own problem. do you think developers should jump through hoops every time a video driver is faulty?

After a couple of months, I expect the developers to fix or at least attempt to mitigate the issue given that their own support says, "we are working on a fix". This implies they believe that they can resolve it one way or another and that it's not nVidia's problem. Had they said, "This is a known nVidia issue that will be addressed in a future version of their driver", then I would have a different expectation. And no, further nVidia updates have not resolved it.

i used to compete online and it is a well known fact that it is not wise to update a video driver just because a new one comes out lol

I built a new system in mid October and that was the driver which was available at the time which was a lot newer than my prior system. I noticed their plugins weren't working, so I opened a support request. I have lots of other plugins that utilize OpenGL and none of them are having issues, so common sense tells me this likely isn't some widespread nVidia issue, rather it's with a framework or implementation (maybe something custom) that Sonible are chosing to use.

In any case, it's not life changing or the end of the world for me -- not really worth any drama. It is what it is and I have plenty of other options to use for EQ and compressors. If I'm being realistic, Sonible is usually my 4th or 5th choice anyway. When they work (as in context for that song/material), they can work very well, but they aren't my primary choice either. Do I think it'll eventually get fixed? Yes, at least I hope so. Does it still impact my decision to buy new plugins like Smart:Limit knowing they aren't particularly fast to address known issues? Absolutely.

[sl1200mk2]

So how does it compare to Elevate, L2, etc?

I haven't had enough time to make an informed opinion about it yet. So far, I've only tried it for a few minutes and just experimenting with it, not using in the same context that I would for an actual project. For me, limiters are very material dependent and they all have their own flavor. My primary choice is generally Pro L2 and Voxengo Elephant is likely second. If someone is wanting a particularly loud or pushed/forward sound I like Elevate a lot for that. My mastering template has a few different limiters loaded up where I can quickly audition them to see what's working well for that given track.