Creating a Playlist with similar volumes.

I have been using Audacity for many years and I finally want to learn what is necessary to create a pleasing Playlist for my listening pleasure. I use only FLAC compressed music which I have ripped from my own CDs. I then want to Normalize/Amplify or whatever such that I can hear all the songs without having to adjust the Volume every time a different artist is randomly chosen on playback. I, of course, have some hard rock and some softer folk music in my repertoire.
I am using Audacity 2.1.3-alpha-Nov 22 2016 on a Linux Fedora 24 Distribution. On my last attempt I setup ‘chains’ and ran them using Normalization. My settings were
—> Normalize ApplyGain=“1” Level=“-0.1000” Remove DcOffset=“1” StereoIndependent=“0”.
The improvement from the raw ripping was a much improved from the original rip but there is still noticeable differences when different genres are playing, which require actual volume adjustments to be used.
I understand that the broadcast industry uses something to make their audio output more consistent, item to item. If normalizing is done one song at a time I understand that the peak for that song is used when doing any adjustment. If many songs are grouped together and normalized I am not quite sure what actually happens. Does it use the individual peaks, the same as doing them one by one, or does it use the highest peak of all the songs (in the group) being normalized?
Any assistance would be welcome…

The best way is to use a player that supports “ReplayGain”. See:
Several audio players for Linux support ReplayGain, including Rhythmbox.

Hi Steve
Thank you for your reply. OK. I had not known that ‘replaygain’ had that capability. I will check to see if the Infotainment system in my Tiguan has that option. I am already using Rhythmbox in house to play my music and I do have ‘replaygain’ as a plugin, which is invoked. I never really noticed that the problem was nowhere near as bad on my in house listening, but I do now.
Can the .FLAC files be made to contain the ‘volume stable’ digital data (values), which would represent the application of the ‘replaygain’ machination directly to an exportable file, whereby the ‘replaygain’ option would not be required to have the desired affect when played on any FLAC capable player? There are a couple of questions, that were in my original question, that I would like to get an answer to.
1.0 If normalizing is done one song at a time I understand that the peak for that specific song is used when doing any adjustment is that correct?
2.0 If many songs are grouped together and normalized as a group (say you select a group of 50 random files) Will all of those files be adjusted relative to a group level or are they essentially still adjusted relative to each specific files own peak?
I thank you for your answer and I will now pursue the ‘replaygain’ as a best solution to my original question. Great work…

When using Normalize in a Chain for batch processing, each processed file is normalized so that its peak level is the specified level.
(in later versions of Audacity, “Chains” have been renamed as “Macros”)