It's indeed compression, but not the mp3 / ogg / whatever compression.

It's the dynamic range compression [1] that makes everything louder, makes everything "pop" more, so there are no nuances left. This is, of course, a mixing / mastering step. But vinyl records of 60s-80s applied it much less than the CD records of 90s-2010s. Those who produce vinyls now still follow the tradition, and the customers expect it.

[1]: https://en.wikipedia.org/wiki/Dynamic_range_compression#Musi...

It got ridiculous in that I would buy vinyl because they were often mixed better, and differently, than CDs that for most of the late 90s-00s were too often a distorted over clipped wall of sound.

I'd then record the vinyl and burn it onto CD for the car. :)

Vinyl has the benefit of a physical limitation, if you add too much loudness the needle will bounce out of the track. You have to master a record in a specific way or it won't work at all.

Maybe the article author is confusing data compression with dynamics compression [0]. The latter is a real problem with the terrible "loudness war" [1] going on for a long time, with many audio pieces (dynamics-)compressed to death. And in that regard, many albums are indeed better on vinyl: not because the format is better than modern digital formats... but because the vinyl version is often mastered separately, with much less dynamic compression than the digital versions. It's really sad that one has to buy the worse format to get the better mastering...

[0] https://en.wikipedia.org/wiki/Dynamic_range_compression

[1] https://en.wikipedia.org/wiki/Loudness_war

No, different master.

A mainstream master, intended for an ordinary consumer, is created with low dynamic range to sound "loud", which is perceived as "better". Also, mastering engineer knows that most of listeners will listen to this on a very shitty speakers and adjusts frequency balance accordingly.

However, when you do a separate vinyl master, you can safely assume that your audience will not compare this version to others on youtube or radio, and that they will have better equipment - so you relax the compression settings, switch off L2 maximizer and allow the track to breathe a little.

Mmkay, so you should just rip the vinyl master once and then use differently/better mastered version digitally.

Yes. Doing that with well mastered 12" singles makes for great digital copies.

The funny thing is vinyl is compressed too. The RIAA curve trades bandwidth in less noticeable sections for bandwidth elsewhere where its better used.

Thus different frequencies have intentionally different noise floors.

The second album I ever bought was Fragile by Yes. There was a deep scratch in the first song on the second side, deep enough to cause skipping and jumping.

Even on the CD version of that album, I still "hear" the scratch. The unscratched perfect version sounds wrong. Can't get it out of my mind.

Have you ever listened to 100 Original Pressings Of The Beatles' White Album Played At Once?

https://www.buzzfeed.com/perpetua/100-original-pressings-of-...

https://soundcloud.com/dustandgrooves/white-album-side-1-x-1...

http://dustandgrooves.com/rutherford-chang-we-buy-white-albu...

The fact that it’s digital, and not analog, means it’s (more) lossy in comparison to the original sound source. It’s not so much the compressed part, but the digital part. If it’s digital, it’s approximating points on the sound waves with “steps” (looks more like stairs than a smooth wave) and missing the nuance between those steps, where as analog, in this case vinyl, is a closer representation of the actual, continuous sound waves.

At least, that’s how I understand it.

That’s a misconception. Monty from xiph.org explains it well. see https://youtu.be/cIQ9IXSUzuM

What a fantastic demo!

This is engineering education of the highest quality.

I was about to take up arms and start explaining Nyquist to the parent comment, but I’m glad I chose to watch this video instead.

Thank you for posting it.

How are the Xiph guys doing? I followed them closely during the early days of their new video codec attempt. I’ve fallen out of the loop since then.

I agree. I had no idea just how many errors I had in my understanding of digital signals.

Excellent, and in my view mandatory viewing for anyone interested in audio and signal processing.

That’s a common explanation, and it’s true at low sampling rates. However, we can fairly provably sample at greater rates than the human ear can perceive. A lot of the actual difference between vinyl and more recent digitized recordings is different decisions made during production, most notably performing volume compression to increase perceived loudness, which definitively reduces the nuance of the final product (but also leads to higher ratings for singles).

You understand it wrong.

Read up on reconstruction filters.