Hands up - I could have done better !

If you saw either of my two videos about inter-sample peaks and True Peak limiters, there's a chance you might have ended up being confused.

This was called out during some interesting conversations over on the Mastering Forum on GearSpace, where several members pointed out that I'd glossed over some important details in those videos - and, I agreed with them.

Specifically, in both videos I used the terms "ISP" or "inter-sample peak" to refer to two different types of peak level, without explaining the difference. They're closely related, but there are some important and interesting differences.

I'm not the only one - many other people also use the terms loosely, but it can lead to some incorrect conclusions, so in this video I've tried to set the record straight, diving into the details and exploring the topic in more depth. Some of the topics include:

• The difference between inter-sample peaks and DSP overshoot, and
• Why it matters
• Why different meters sometimes give different True Peak readings
• What over-sampling has to do with it all
• The best way to test the "sound" of a True Peak limiter, and
• When a True Peak limiter may not do what you were expecting

Hopefully you'll find it interesting - one of the people I was talking to on GearSpace was Alexey Lukin, the principal DSP developer from iZotope, who really knows his stuff ! As a result I ended up talking to him on my podcast about all this, and you can hear my interview with him here.

It's a question I get asked fairly often - "if YouTube reduces the loudness of loud songs, why don't I always hear it ?"

And the answer is often that the person is listening to YouTube Music, rather than the main video platform. YouTube Music is a dedicated audio-only subscription platform, with only a tiny fraction of the number of users of the main video platform. But many people don't realise they're separate sites - it's certainly not immediately obvious to most of us, despite having a different URL and logo.

OK, so why do we notice loudness normalisation on YouTube's video platform, but not YouTube Music ? My answer used to be that YouTube Music simply doesn't use loudness normalisation yet, but in fact I've recently realised that's not correct. The platform IS reducing the loudness of the loudest songs, just not nearly as much as YouTube Video - because it has a different Distribution Loudness level.

Miley Cyrus has a huge hit with “Flowers” right now, and I recently shared two different versions of her song “Jaded” on social media, asking people which version they preferred, and why.

If you saw the post and wondered why I shared it, or if you missed it and would like to take a listen for yourself, or if you're just curious why I care, check out the video above.

And let me know which one you prefer (and why) in the comments on YouTube !

People tell me all the time that True-Peak limiters sound worse than older sample limiters - that they "kill transients".

That hasn't been my experience, but I think I know why people think that - it's because often the way they're testing these limiters may be deeply misleading.

In this video I demonstrate how and why this can happen, and suggest a better, more objective way to compare sample limiting and True Peak limiting. The results may surprise you !

(If you're not 100% sure what an inter-sample peak is, or whether they might ever cause an audible problem in your audio, I made another video to demonstrate - click here to check it out)