I wish it was still being actively developed.
Full Version: Discussion of FLAC's comparison table
It looks like LA is still the compression king (at expense of CPU cycles) overall. I wish it was still being actively developed.
I wish it was still being actively developed.
QUOTE(vinnie97 @ Feb 10 2007, 17:57) 
It looks like LA is still the compression king (at expense of CPU cycles) overall.
I wish it was still being actively developed.I doubt that this is true if you compare it with OptimFROG 4.600ex (Experimental). I have seen very much data and performed some own tests and i am quite sure, that it will be the king, when a stable version has been released...
I await that day with baited breath! 
QUOTE(TBeck @ Feb 10 2007, 23:16)
QUOTE(vinnie97 @ Feb 10 2007, 17:57)
It looks like LA is still the compression king (at expense of CPU cycles) overall.
I wish it was still being actively developed.I doubt that this is true if you compare it with OptimFROG 4.600ex (Experimental). I have seen very much data and performed some own tests and i am quite sure, that it will be the king, when a stable version has been released...
In my limited tests, i found that LA often reached a lower size than Optim Experimental, even though Optim took a longer time. However, they both took an INSANELY long amount of time.
QUOTE(kanak @ Feb 10 2007, 20:12)
In my limited tests, i found that LA often reached a lower size than Optim Experimental, even though Optim took a longer time. However, they both took an INSANELY long amount of time.
As always "limited tests" is very important. Lossless audio codecs have different weaknesses and strengths. Example: while TAK on average should be on par with Monkey's High, some tests show it performing only at Monkey's normal level, others even better the Monkey's Extra High... It's very difficult to create a representative test set. For example it is definitely not sufficient to select a couple of files from each genre.
Some of the data i have seen comes from a test performed on more than 50 GB of audio. I tend to trust it, but the result is also not the final truth...
QUOTE(TBeck @ Feb 11 2007, 01:29)
QUOTE(kanak @ Feb 10 2007, 20:12)
In my limited tests, i found that LA often reached a lower size than Optim Experimental, even though Optim took a longer time. However, they both took an INSANELY long amount of time.
As always "limited tests" is very important. Lossless audio codecs have different weaknesses and strengths. Example: while TAK on average should be on par with Monkey's High, some tests show it performing only at Monkey's normal level, others even better the Monkey's Extra High... It's very difficult to create a representative test set. For example it is definitely not sufficient to select a couple of files from each genre.
Some of the data i have seen comes from a test performed on more than 50 GB of audio. I tend to trust it, but the result is also not the final truth...
Tbeck, could you give a little more details? like how much better Optim was for example. I'm really curious.
QUOTE(kanak @ Feb 11 2007, 06:17)
QUOTE(TBeck @ Feb 11 2007, 01:29)
QUOTE(kanak @ Feb 10 2007, 20:12)
In my limited tests, i found that LA often reached a lower size than Optim Experimental, even though Optim took a longer time. However, they both took an INSANELY long amount of time.
As always "limited tests" is very important. Lossless audio codecs have different weaknesses and strengths. Example: while TAK on average should be on par with Monkey's High, some tests show it performing only at Monkey's normal level, others even better the Monkey's Extra High... It's very difficult to create a representative test set. For example it is definitely not sufficient to select a couple of files from each genre.
Some of the data i have seen comes from a test performed on more than 50 GB of audio. I tend to trust it, but the result is also not the final truth...
Tbeck, could you give a little more details? like how much better Optim was for example. I'm really curious.
Sorry, i am not allowed to provide details, before the author of the big test has published his results.
I want to add something to my previous post: While i think it is important to insist on the use of representative file sets, if a tester publishes a comparison of the general compression efficiency of different codecs, this is less important for specific personal use. Possibly a codec on average isn't the best, but is the best for a specific genre (for instance classical music). Then if most of your music is from this genre, this codec could be your best choice despite its weaknesses when compressing other genres (IMHO).
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.