I've been playing with getting the bitrate column in my fcs to display a different background depending on the bitrate of the corresponding file. The code I've come up with so far is:



As you can see, the problem arises when trying to compare the bitrates of MP3's and the lossless files in my library such as FLAC and APE. This is why I have seperated the two types of codecs in the above code. If the baseline value for all codecs is set to high (to account for the lossless files), then the color difference between the MP3's is negligible. Too low and the high bitrate lossless files will not be included in the output.

I was wondering if someone could help me tune this to get a relative ratio going for the lossless and MP3 files so that the MP3's @ <128 would be red, and the best FLAC files would be green.

I think the blend command works the best, as I have tried to code for each type of mp3 by using the $greater command and choosing bitrates at 128, 160, 192, 256, 320, and 321, which gives pretty good results, but can't differentiate between a 192 kbps CBR MP3 and a 161 kbps VBR MP3.

Any help would be appreciated.

Maybe the above screenshot will better show you what I'm trying to accomplish. Eventually I want to have a completely color coded method of interpreting music metadata.

The RG and AG signify the presence of Replaygain and Album gain, respectively. Eventually they will be given their own columns using a solid green backgroud as indicating its presence, and a solid red background indicating is absence.

Basically, I want to be able to go through my collection, looking for any red, which will clue me into problems that I have with certain tracks.

I also want to design something more visually pleasing and that contains more information in the color when indicating bitrate (quality). It is for this reason that I do not like to use set values to determine the color that is displayed (i.e >128, >160, >192). It is for this reason that I chose the blend command.

The problem arises when dealing with a data set with a large standard deviation. It is for this reason that lossy and lossless encoders must be treated differently (as explained in my post above).

As in the example of differentiating VBR and CBR MP3's, obviosuly the former will be of somewhat higher quality. How much so is what I am trying to get some input on. Another problem is determing the comparison value of a 320 kbps lossy file to a lossless rip. Obviously it will not differ as much as say a 64 kbps MP3 to a 128 kbps MP3. The quality gained by using higher bitrates seems to be hyperbolic. This is a problem because Foobar, as of yet, does not support exponential functions. Even if it did, assigning a value to this would be quite a challenge.

What I need is someone to help determine a value for determing the relative quality of a song by its bitrate. Since logarithmic functions aren't at our disposal, the idea of comparing FLAC to MP3 is out of the question. Even when separating the two formats, as I have done in the code I posted above, a non-linear change in quality occurs within a specific encoder when moving to the higher bitrates.

What I have determined is that it will take an integration of some sort to determine these values (which may indeed be quantitative (unless there is documentation already on percieved quality differences and certain bitrates).
i.e.
blend color a to b for bitrates between c and d
blend color b to e for bitrates between d and g
blend color e to h for bitrates between g and i

...and so on and so forth.

This is what I need help determing, not coding as the above responses indicated.

EDIT: Updated Foobar. Green column is presence of RG. Far Right Column is Quality. Color of Track is Playcount. Purple Heart is presence of Ratings tag.