Interesting controlled test from the BBC

[Pim]

Well it’s always important to realize the motivation behind a study and their indented outcome.
In this case BBC wanted to showcase that their broadcasting with reduced / compressed file size is obviously not making a difference. Well they achieved their desired result ;-)

The only broadcaster to ever go out of their way to design a speaker so their recordings are more accurate. Enough said.

[Confused]

I found the BBC paper to be very interesting. One point I would pick up on was that the test criteria was based on picking up "artifacts", rather than a subjective view of the sound quality.

The reason that I mention this is that I think that I can just about pick up the differences between say a 320kbps file and a 16/44.1 version. Although I do not hear "artifacts". It is just that the 320kbps version is maybe a touch "softer", with a fractional loss of detail. Curiously, with some lesser mastered recordings this might not be a bad thing, as it fractionally takes the edge off harsher recordings, and might make the specific recording a little less fatiguing to listen to, even if at the borderlines of sound quality something is missing. The BBC criteria lists the "impairment" as "Annoying", "slightly annoying" and so on. To my ears, something like AAC 320kbps may lose something, but it is not annoying at all in terms of artifacts. So I do wonder if a very slightly different result may have be obtained with a different criteria. A minor point though, I think the ultimate conclusion is valid.

I think formats like AAC 320kbps in absolute terms can sound superb. Certainly I would much rather listen to something with decent mastering at 320kbps than a lesser mastering at 16/44.1. It is perhaps also worth remembering that modern compressed formats like AAC and Ogg/ Vorbis are vastly better than early MP3 formats, but I suspect the legacy of MP3 may be unfairly damaging the reputation of AAC and similar.

It is perhaps worth mentioning that I cannot really tell any difference between 16/44.1 and higher resolution formats, if from identical masters. I proved this by trying the Mark Waldrep / Dr. AIX blind test, my final results we no better than if I had picked at random.

So for me AAC 320kbps is absolutely fine and very nearly as good as 16/44.1, and 16/44.1 is more or less as good as anything.

I am sure some others will disagree!

One last point to make, I often listen to the BBC and I can easily get the AAC 320kbps links in Roon. Technical debates aside, I find this to be most enjoyable.