Could you post signal path screenshots of both ‘ok’ and ‘problem’ tracks playing?
I think that’s what I’d expect, so not clear why it sounds ‘not ok’.
What are the tracks? If I can find them on Tidal I’ll see if I get the same numbers - with the obvious caveat that I could be listening to a different master…
“Not OK” - have to turn the level up. Steely Dan “My Old School”
“OK” - Aes Dana, album “Inks”
Torben
1 Like
Here a small comparison on a low volume album: “Autumn Rain In A Secret Garden” - Csaba Pengő
Foobar2000 - + 2.4 dB
and Roon - + 0.0 dB
That would be a different of 2.4 dB
Torben
Is Foobar using R128 or ‘classic’? What target is set?
(I’m going to find my laptop - and see what version of Foobar I’ve got!)
foobar2000 is using R128 LUFS -18. And works nicely. @Torben_Rick you should be able to scan and add RG tags to your 600 local albums in just a few hours in a batch mode with only a few clicks of your mouse to get it started.
Keep in mind that adding the RG tags won’t help you if you do not tick the option to let the RG tags override Roon volume adjustment tags. As you’ve shown above, Roon in your example would play the album at a much lower volume than use of the RG tags (+0 in roon vs. +2.4 in foobar RG tags).
I’ve updated Foobar… so with luck I’m now looking at the same options as you… which deflaults to EBU R128.
This is my result with what I’d consider a faily typical CD rip:
All the track gains are negative, which I think is what I’d expect from a remastered compilation from 1993.
I think what you are seeing is Foobar indicating that a positive gain is needed to bring the album up to -18LUFS; however, because the peak level of one track is particularly high, Roon can’t apply that much gain without clipping.
TL;DR - try setting Roon levelling to -23LUFS and see what happens…
@garym am I right in thinking that - when using RG tags - Roon will still back off positive gain if it thinks the peaks will clip?
@AndyR I believe that is correct. But I’ve not tried to test that.
1 Like
This is an R128 analysis of My Old School…
… which to my mind confirms the -2.5dB RG track gain is about right.
@Torben_Rick - what does Foobar show for your file of My Old School?
Is that from a download of Countdown to Ecstasy?
Here a another example: Peo Alfonsi & Salvatore Maiore - Alma
Roon adds +1.2 dB
Foobar2000 adds + 5.23 dB
Foobar2000 would probably the right one compared to other albums.
Torben
OK. Just for my benefit then… set the Roon target to -23LUFS and post the result. 
This is -14LUFS
Peo Alfonsi & Salvatore Maiore - Alma
and this is -23LUFS
Peo Alfonsi & Salvatore Maiore - Alma
Torben
And now, does the different target value change the relative loudness perception between different albums or tracks to the better?
It’s about perceived peaks in loudness, due to compression used in more modern recordings they will still sound louder as the dynamic range isn’t there, where older recordings it is. Replay gain/volume levelling can’t account for the more modern compressed mastering. It can help it but not negate it completely. There will always be differences that’s why they compressed music to start with so it sounded louder on radio. Loud sells apparently.
2 Likes
I know but by default it doesn’t. Just simplifying it for him as no need to add them purely for Roon. it won’t make much difference letting Roon use its own analysis over embedded RG tags if done to R128 which all the apps do these days.
Sure, but that the smallest problem. Many of the albums that I have, have been released the last couple of year. Peo Alfonsi & Salvatore Maiore - Alma is good example. But there are many more
Torben