ISP/True Peak limiter test

ISP, Inter Sample Peak or TP, True Peak, refers to peaks in the analog domain. That is, how your peaks will look after your waveform has been converted from a digital stream within your computer, phone, iPod or CD to electrical impulses that can be amplified and played back through your speakers.

Aren’t the waveforms the same as they appear on my screen after the D/A conversion?
No, Your D/A will create a waveform that continuously moves from one sample point to the next. In that process the newly created waveform may peak above the 0dBFS ceiling which individual samples adhere to, thereby causing distortion. Just how bad this distortion is depends on the converter used. In some extreme cases, the actual peak can be as much as 3-4dB above the sample peaks detected in the digital domain.

 

Many of the top mastering business don’t care about ISP so why should I?
No, you don’t have to treat or be aware of ISP but:

  1. Digital limiters without oversampling (or other ways of detect ISP) will react to the digital waveform, not the true waveform, and thus won’t react as a limiter should react. This does not mean it will sound bad in any way, just that it don’t react as intended.
  2. -Tip: If you want to work with limiters that lack oversampling;  up-sample your 44.1 or 48KHz projects to a higher sample rate and by that have the limiter react more accurately.
  3. We’ll most probably have a loudness standard for streaming media within a couple of years and that standard will be ISP aware and set at -1dBTP. This means that if your audio peak at +2dBTP it will be turned down by at least -3dB before reaching the end user. All that extra loudness you gained by letting random equipment create a positive peak (that might distort) will be lost so it will be meaningless.
  4. Encoding to lossy formats; ISP will make it harder for encoders to do a good job. If you don’t have the tools or knowledge to check how your audio will perform post encoding I would recommend to stay away from positive True Peaks.

Since it’s more or less guess-work, or at least really hard to create a real-time limiter that handles ISP perfectly, we thought we’d put as many limiters as possible to the test. So a handful of mastering colleagues and I (Ian Stewart, Sigurdór Guðmundsson and Johan Eckerblad) went to work.

 

The Test

We used a mastered track that we boosted by 7.5dB (the peaks were at -1dBTP/dBFS in the mastered file so max Gain Reduction would be 6.5dB), we set the limiter ceiling to -1.0dBFS (or -1dBTP if available in the plugin), rendered it as 44.1KHz, 32bfp wav and measured the True Peaks of the resulting waveforms. Limiters with a result as close to -1.0 dBTP as possible have handled ISPs the best, whereas values over -0.8 dBTP (shown with increasing values in white, through yellow, to red when they exceed 0 dBTP) mean that the limiter has failed to handle ISPs to within the margin of error for measurement we observed during testing (more on that in a bit).

Here’s the track that we used:

https://siggidori.bandcamp.com/track/morbit-orbit

 

  • “Yes” or “No” in the TRUE PEAK column indicates whether the developer mentions or claims that the plugin handles True Peak or ISP, either in the manual or marketing.  A “No” does not necessarily mean that it doesn’t handle or try to handle ISP, just that it’s not explicitly stated by the developer.
  • The results published here are measurements taken using Izotope RX. Additionally, we did measure with Sequoia, Wavelab, and Nugen, and even though we observed slightly different results, they were negligible. The biggest differences were in iZotope’s products.  Since they probably use the same detection algorithms in all their software, this was expected. The other software measured within a 0.1 to 0.2 dB difference, which still puts iZotope amongst the best performers.
  • Where other settings such as Attack and Release were present we used either the default or adjusted it to a value as close to standard as possible
  • This test did not take into account sonic qualities at all.  That will have to wait for our planned podcast or another post.
  • It’s not recommended to use any of the limiters with a positive score in 44.1KHz projects. They should handle ISP better as your sample rate goes up.
  • If your favorite Mastering Limiter ain’t here, send us a link so that we can try it out. However, we’re not in the business of buying every limiter out there, so any you’d like us to try must have fully working demo-versions.

96 Comments so far:

  1. You didn’t test Avid’s Pro Limiter which is supposed to be a true peak limiter. As far as I know Dyn3 is not true peak.

  2. I second the Flux products. The Pure Limiter I would love to see the results. Thanks for all your work guys!

      • The free Limited-Z doesn’t claim to address ISPs. So why would you include it in a test of true-peak limiters. In fact there appear to be a number of limiters tested here that are not true-peak or ISP limiters. So readers, please keep this in mind that even though this article is about true-peak/ISP limiters the author has decided to include limiters that are not true-peak/ISP – this should not reflect negatively on these limiters as they do not claim to support this functionality.

        • If you take a more closely look at the list you’ll see that we clearly state which limiters are claimed by their developers to handle ISP or not. We added these limiters since many of them do handle ISP even though the developer doesn’t mention anything about it in their marketing. A digital limiter needs to address ISP to be able to do what it is supposed to do, if the signal goes above the set ceiling in the analog domain it has failed its task to limit the signal. The natural approach while coding a limiter would be to always use oversampling or other methods of TP detection which most developers do. It would be unfair not to include these limiters.

        • Another reason why we did it (compare “normal” limiters to True Peak ones) was to get a comparison of how they react. So perhaps it should be called a “Limiter test with special interest on ISP limiters”, sexy, I know.

  3. Very interesting comparison!

    Here are some suggestions of Limiters to (maybe) include from my side (if possible):

    1.) The built-in Limiter from the freeware “Audacity” – http://www.audacityteam.org/

    … and (since the IK Multimedia “Stealth” Limiter was already listed here) …

    2.) IK Multimedia’s Brickwall Limiter – http://www.ikmultimedia.com/products/trbricklim/
    3.) IK Multimedia’s Classic Limiter – http://www.ikmultimedia.com/products/trclasslim/
    4.) IK Multimedia’s Vintage Tube Limiter – http://www.ikmultimedia.com/products/trvintubcomplim
    5.) IK Multimedia’s Precision Limiter – http://www.ikmultimedia.com/products/trpreccomplim/

    Thank you for considering! – 🙂

      • NOO, It’s not true. Sonnox Limiter has TP in all format AU VST RTAS AAX.. You can realize a ISP/TP Limiting in this way: turn on the “Auto Comp” Button, then the “Recon Meter” to see all the TruePeak Corrected.
        Please do not confuse the information spreading misconceptions.
        Also, ISPs should only be considered inherent in processors that deal with standardization and Loudness maximisation.
        It’s not possible to make a ISP comparing between FabFilterMB and FabFilter Pro-L. They Are two things working in two completely different ways even if you activate the maximum value of Lookahead on MB.

        • I don’t agree that ISP should only be considered in standardization and Loudness maximization, the problem occur because 44.1KHz ain’t a sampling frequency high enough to accurately draw an exact copy of the DA converted waveform so the limiter will react faulty making it more of an Dynamic shaping tool than an process to stop peaks. One remedy is to work on higher sample rates, you can both hear and measure that even 88.2 make limiters behave more accurately. At 192 most limiters, even the ones that are not ISP aware, handle ISP perfectly. This test is for projects at 44.1KHz.

          About FabFilter; yes, agreed, you can’t compare them as similar processors BUT you can handle ISP with both. We chose to include Pro MB (which is not a limiter) to make people aware that you can adress ISP and limiting in general with other means than a dedicated limiter.

          • You’re right, and you’re wrong… Sure, a 100:1 ratio is pretty much a brick wall but when you design a limiter you put your efforts on slightly different characteristics than with a compressor.

          • Of course 🙂 I would indeed expect plugin developers to make their limiters “limit” properly 🙂 But a limiting ratio is a limiting ratio 🙂 … and there is even a range to that if you look at analog limiters or emulations.

    • We’ll see what we can do!

      Side note about requested limiters that are not added yet. Some developers make it so hard to install demos that we don’t simply have the time to add them, but we’ll hopefully manage to add them eventually. To perform the actual test takes about 10 minutes, to download and install a demo can take anything from a minute to a day (if you count the time waiting for email confirmation and/or unlock codes).

  4. Thank you very much for your effort going into this valuable comparison of limiters!
    I have compared quite a number of different limiters, and from the point of view of the sound I always felt Kuassa Kratos to be superior.
    But concerning ISP capabilities, I haven’t checked Kratos so far. Current version is Kratos 2 which claims to detect ISPs (which I verified in Kratos’ manual).

    Download with unrestricted demo mode (with noises inserted every now and then): http://www.kuassa.com/downloads/

    Thanks again, best,
    Tobias

      • Thanks a lot for having checked Kratos 2 !

        Just one correction: “No” in the “TRUE PEAK” column is not correct.
        As I mentioned in my previous post, the manual claims that Kratos detects ISPs.
        Quoting the manual, page 2 (“Overview”): “Presenting the latest version of… Kratos 2 Maximizer; featuring …, inter-sample peak detection, …

        Best wishes, and thanks again,
        Tobias

    • We’re not going to include UAD LA-2A, it’s not that kind of a limiter and you can’t set it according to the specs for this test so the result wouldn’t be comparable.

      • You can demo it for 14 days. Here is a copy paste from their website:

        This audio processor module is part of T-RackS CS “Custom Shop” Mixing and Mastering Plug-In Collection. With T-RackS CS you can:

        -install all of the available modules at once
        -directly access Custom Shop online store from the module for immediate trial or purchase
        -try out any processor for a full 14 day trial period
        -purchase any module with credits (via Custom Shop) or currency (via Custom Shop or IK online store)

        • Stealth Limiter added… It did Ok, I tried a couple of different modes beyond the posted result and it generated similar results.

  5. Hi!

    Now, that you have IK Multimedia’s Custom Shop installed for the Stealth Limiter Test, I wonder if you might want to check out the other IK Limiters I listed here before – here we go again:

    1.) IK Multimedia’s Brickwall Limiter – http://www.ikmultimedia.com/products/trbricklim/
    2.) IK Multimedia’s Classic Limiter – http://www.ikmultimedia.com/products/trclasslim/
    3.) IK Multimedia’s Vintage Tube Limiter – http://www.ikmultimedia.com/products/trvintubcomplim
    4.) IK Multimedia’s Precision Limiter – http://www.ikmultimedia.com/products/trpreccomplim/

    Would love to see a comparison for those as well . – Thank you for considering! – 🙂

  6. What a great and interesting test – Hat off for the strong achievement!

    Would like to see onboard/stock Limiters from other DAWs like
    Apple Logic, Steinberg Cubase, Cakewalk Sonar, Avid Pro Tools.
    Any chance you could add this?

    of course you cannot have any DAW (purchase of Dongles etc.)
    but maybe you could even ask people who have theses DAWs e.g.

    that would be totally great!

    best/kind regards
    .

  7. Hi!
    Did you already measure Avid Pro Limiter? I can test it for the list (both Native and DSP modes if needed), if you submit the same test material and any instructions you have used?

  8. Please, please, remove the video playing in the background or add a way to stop it. It’s super distracting when trying to read the text.

    • We won’t remove the video but there’s a ton of video blocking add-ons for most browsers so the solution is right at your fingertips. =)

      • I guess you don’t care, but I wonder how many people turn around and leave without complaining because of the distraction. I wonder why you thought it was a good idea in the first place, though. Especially to use video material with a bunch of fast panning. Even just having the same video but with static scenes would be more acceptable. Why pull focus to the sides of the screen when your content is in the center? It’s bad UX and typical careless modern web design. Your loss, I guess.

        • I do care but don’t have the time to work with the site atm so a solution at your end is better. I’m building a new mastering studio and I’m putting all my energy there, site maintenance will be on hold until it’s finished. Thanks for all your suggestions, they’re appreciated.

  9. In Voxengo Elephant settings you can change the default ‘Auto’ oversampling amount.
    When you first install Elephant, ‘Auto’ will be set to 1x.
    Go to SETTINGS > GLOBAL VOXENGO SOFTWARE SETTINGS > Auto Oversampling Level = 8x
    After that ‘Auto’ will always be 8x oversampling for ALL Voxengo plugs you have set to ‘Auto’ in the current session, and in any Voxengo plugin in future sessions..
    I’ve used it for over 10 years and still do 🙂

    • We’ve been busy lately but we have a list with limiters to add and Maximal have been added to that list. It’ll be a while though, sorry about that.

      • No worries on being busy. I appreciate that you have added Maximal to the list! Thank you for compiling this list for the audio community. Cheers!

  10. I am glad to see this post and list being updated! I am also glad you evaluated Elevate – this one was a huge dissapointment because it’s sold as a mastering limiter. As a mastering limiter it should be stellar on handling ISPs. This was the first thing I tested and was disappointed and confused to see it was not. This gives me a sense of credibility of this company.

    I wonder if we can dig a little deeper into the ISP side of things: seems to me that the ISP process involves some kind of reconstruction of the waveform that gets clipped. Would there not be differeant methods to do this and that those methods may sound different? Is there an objective way to evaluate a limters ISP quality? And, would it matter?

    • The best remedy to ISP is to work at a higher samplerate and/or use oversampling, if available. You can also put a gain post your limiter and back off about -1dB to be sure (Sequoia use this method in its built-in limiter amongst others).

      • That’s all true. However there are limters out there that truly handle ISPs and handle them very well. ISL v2 is an example. See my reply to your comment on LVC below…

        Personally, I’d rather use a limiter that truly handles ISP meaning, they do not allow ISPs to occur in their output. Otherwise I’d have to play through the track from start to finish and monitor in an ISP tool like AppleRoundTrip. I’d rather just put trust in a true ISP limiter so I don’t have to manage all this myself. And since there are some good true ISP limiters out there it makes sense to use them and not those that do not handle ISPs very well.

  11. A note about LVC: I saw the entry here for Limited-Z. I have Limited-MAX and just tested it. I didn’t use your methodology. Instead I just wanted to confirm/deny if it was a true-peak ISP limiter. It is not. First, I set it up as a mastering peak limiter with 0.1ms attack, 50ms release, -0.7dB ceiling. I turned ISP on and oversample to 8x and DC filtering. With a target -16 LUFS loudness it only generated an ISP every 60 seconds or so. If the loduness was raised to -14 LUFS it would throw ISPs regularily. I could see using this for -16LUFS as it’s good enough. But no it’s not true ISP. I don’t know what the algo is in the limiter, but I am thinking he decided not to implement a true-ISP correction but rather to make the limiter as good as posible at around a resonable level of loudness. In anycase, Limited-MAX is not a true ISP limiter.

    • What I can remember LVC claim that Limited-Max do handle ISP and that’s why we put that in the comment field. That does not mean it actually does handle ISP but rather that the developer tried to handle them. I developed limiters myself and ISP is a tricky business!

      • Limited-MAX does have an ISP mode. However it still not a true ISP limiter. I am not sure what ISP mode does but it helps to quell ISPs. It has to be used with over-sampling to get better ISP handling. Limited-MAX has over-sampling up to 8x. So it is possible with ISP on and 8x over-sample (at 44.1k project rate) to get Limited-MAX to reliably quell *most* ISPs. But again, unless I really want the sound/feature of Limited-MAX I’d rather just use a true ISP limiter like ISL v2 so I don’t have to mess with these parameter and use hope instead of fact.

        I can imagine that preventing ISPs is difficult. Although I don’t code and do not do DSP work, I can understand how even just limiting would be a challenge. In both cases you are trying to re-construct a clipped waveform. I have no idea how that’s done. and then with ISPs, there is some trickery going on there to predict when an inter-sample will clip. Amazing stuff you guys pull off with these tools!

Leave a Reply

Your email address will not be published. Required fields are marked *

Posted by: Saintpid on