Is this sourced from MP3? Spectral view indicates that it is....

I don't have that show handy, can you post a picture of the spectral analysis? could just be 32k dat, since it was a d7....

but i got that disc from a very reputable trader.

since there's nothing above 16KHz, i'd guess it was just a 32KHz master. there weren't any glitches between tracks were there?

The reason I ask is because I don't want to distro an EQ'ed version if it's mp3 sourced.

Speaking of which, let me know what you all think: http://www.stonepool.com/chrishickman/hailhail.ogg

Ogg Vorbis media file. Winamp plays it either by default or by plugin, depending on what version you have.

Ok well this is sort of from my other thread, but what is meant by "EQ'ed mp3 source" I think i know but i want to make sure.

Let me ask this. If VBR is made to encode a MP3 will it still have the same cutoff at 16K? What would a mp3 file look like that is encoded at 320KBS. Are they all cutoff at 16K still?

I can understand the generational loss over several dozen reencodes. But really what is lost if i encode a file from a source wav to a VBR mp3 or 320 KBS.

Im not trying to go against any grain or create a monster with this thread, im really just trying to educate myself

Youkan

different encoders do different things to the signal before trying to compress it.

seeing a spectral view with no frequencies above 15KHz and some weird looking blocks used to be a dead giveaway. this is just because the way mp3 works at lower bitrates. the higher frequencies are harder to accurately reproduce with compression, and they're harder to hear anyways. so a lot of encoders (and codecs in general) just throw everything away over a certain cutoff frequency, because it would require a lot of bits to compress it well.

when you use lame/vbr, there's a transition band, and it's usually higher with higher bitrates. you might see a tapering off up near 19KHz.

When you've got a recording on cd, for example, that recording is sampled at 44KHz. this means it cannot reproduce any frequencies higher than 22KHz. So when you see something on CD that doesn't have anything between 16 and 22, you know that at some point, it was either filtered or sampled at lower than 44KHz.

so when you see that line at 16KHz, it's harder to tell if it was just recorded originally on a 32khz dat, or if it was compressed somewhere. the d7 only records at 32khz and 48khz. it's certainly possible it was recorded at 32khz. plus, like i said, i got this disc from a good trader who i'm sure would check to see if it was mp3 sourced. if the tracks are seamless, i wouldn't give it another thought.

When you see a cutoff, that usually means that information about the recording was discarded. encoding at 320kbps will give you a compressed file that sounds very much like the original, but if you really pay attention to high frequencies (like vocals, guitar riffs/leads, and cymbals), you'll notice a difference.

from the sound of things, twoheadedboy is doing (or has done) an EQ'ed version of this show. EQ just makes some frequencies louder and others softer. if there's too much bass (low frequency) in the recording, you can reduce that with an EQ. If the recording sounds muffled, maybe the highs need to be brought up. EQ can't fix everything (can't bring back the information that was just plain thrown away), but it can improve a lot of recordings.

and checking to see if it was mp3 sourced, well he was just making sure there wasn't a better copy that he should be working with. no sense on working with garbage if there's a better copy available.

i am starting to see things clearly now. You really have given me some good knowledge to go on my friend. I appreciate it muchly..

Youkan