[Polygon]

We talk a lot about the Mp3 vs CD, and how most think that Mp3s have quite a bit of sound degridation even at 128kbits 44,100Hz. Well what I am going to do is post some Mp3s that I have done myself from CDs that I do own. I want you to tell me what you think of the sound quality and I want to know if you can guess what codec was used for their compression.

Song 1.

Song 2.

Song 3.

 

[MTRH]

what is your goal here?
what codec performs best at the horrible 128k bitrate?

 

[Polygon]

Well, in eccense, yes. Plus I just want your opinons on how you think these three Mp3s sound.

 

[ff123]

Sample 2 (Rock With You) sounds bad between 15 and 19 seconds. Also the cymbals sound bad starting at about 2:06. Without the original to compare with, though, I can't say for sure that it's because of the compression, but that would be my guess because it has the typical mp3 sound to it.

I think I hear some watery sounds now and then in Sample 1 (Superman). But otherwise, I'm not too bothered by it.

Sample 3 (Tuna Too Much) sounds fine to me.

Some people are bothered by mp3 artifacting much more than I am, because most of the time with a decent mp3 encoder, I am happy to listen to 128 kbit/s files.

Hint: You should remove the ID3 tags to make it slightly harder for people to guess the codec, although that won't stop people using EncSpot.

ff123

 

[Polygon]

Hehe, yeah I should have, but it only lists the encoder I use. I don't even know which codec it is using to encode. Thank you for the input.

 

[Magicthyse]

Well, I'm by no means an expert and I didn't look at the ID3 tags as I just clicked and listened but I didn't like 2 at all.

PS. Listened on a Grado SR60 off a SBLive! Platinum.

 

[chych]

I think it would be rather interesting for us to have two songs available, in WAV format, one that was ripped as mp3 at 192kbps LAME (or something decent like that) and one the original, and see what we thought...

 

[ff123]

Quote:

Originally posted by chych I think it would be rather interesting for us to have two songs available, in WAV format, one that was ripped as mp3 at 192kbps LAME (or something decent like that) and one the original, and see what we thought...

It doesn't have to be in wav format, which would be a huge download. A good test can be set up quite easily using just the mp3 files. If one of the encoders you want to compare is Lame, you can even dispense with the lame mp3! The trick is to include the lame command line binary encoder/decoder with the test sample download. For an example of such a test, see the one I conducted in this thread:

http://www.hydrogenaudio.org/forums/...&threadid=2075

This was a test of an audio comparison utility I'm currently writing (my very first C++ program and my first win32 program to boot). Three encoders were compared using a 6 second sample, and the download was a trim 1.3 MB.

ff123

 

[penvzila]

Ogg Vorbis is pretty damn good. It is variable bitrate, and the encoding options are kind of like a scale from 1 to 10 (i think) incremented by .01 or .001 (cant remember exact.). The newest version of the encoder is excellent, fixing some glaring quality issues. Plus ogg is completely patent free and open source.