Not the kind of thing you people would probably want to see in someones first post.. but

I got a burnt cd from a friend.. and I'm pretty sure it's been mp3 sourced. Spectral view in EAC's process wave thing shows a band at about 15khz, below the line is mainly bright orange/red.. above the line is just dotted with blue. That's the basis for my assumption.

I ususally use alt-preset-standard to encode, but that would be redundant for something like this. I'm looking for a setting for it, I just don't know the names of anything below alt-preset-medium.

So does a 15khz cutoff rougly equate to a particular bitrate, so I can base an encoder setting upon it? I'm not too fussed that it's done badly, and I definitely won't pass it on to anyone else. I just want to put it on my computer without encoding a ton of redundancy into it.

well redundant it may be -- but re-encoding from original to wav to mp3 to cd-r to wav to mp3 is redundant as well -- face it! The easy and n00b way out of this (although not always 100% easy to tell) is to --alt-preset standard either a few tracks or the whole album, EncSpot scan it , find out what the avg bitrate seems to be for most tracks, and just re-rip the cd in that CBR bitrate. again, no matter what you do the whole scenario is redundant. I hate re-encoding, but sometimes it must be done good luck

It's very simple:
--preset extreme = -V0
--preset standard = -V2
--preset medium = -V4

If you want something inferior to medium preset or -V4, use -V5, -V6, -V7....
If the signal doesn't go beyond 15 KHz, it might be interesting to resample the files by using the --resample 32000 command.

Most are ~160 kbps, except for one which is 202kpbs...

so.. vbr with a cap of 160 or 192 and maybe -v5
does that seem sufficient given the circumstances?

Yeah sure go for it . . . but trust me , you will feel icky for attempting to do a re-encode . . . it comes back to bite you in the ass . . . one day you will purchase the original and feel catharsis whilst listening to your high-compression musik

ok, cheers to both of you.

Do you mean a more or less constant line around 15 to 16 kHz? This would look more like electromagnetic radiation of a video monitor captured by the recording equipment.

Nowadays many sound engineers seem to be too lazy/stupid to avoid that.

nah, the line is a distinct cutoff like this:



this is not my file/pic, I found it elsewhere on the forum (thanks to BadReligionPR), but it's very similar.
The only differenec is where its black in this image, I still get some sparse blue bits.

Where is your cutoff frequency ? Is it 14kHz as in these pictures ?

Capping is not a good idea, let the encoder decide how many bits it needs.
Just follow guruboolez's advice - he really knows what he is talking about

Can you hear anything that sounds like encoder artifacts from your CD?

Try TauAnalyzer or auCDtect to analyze its source.

For VBR, -V 7 has a lowpass transition band at 14581 Hz - 14968 Hz
For ABR, average bitrates from 88 to 103 have a lowpass transition band at 15115 Hz - 15648 Hz
For CBR, 96kbps has a lowpass transition band at 15115 Hz - 15648 Hz

Integrating:

auCDtect gives them a 95% chance of being mpegs.

Sunhillow: Wouldn't a non-capped encode be equivalent to transcoding from a lower to higher bitrate? Or is there some possibility of audio that can be "salvaged" by not having a cap?

If not, in the case that it's 96kbps CBR (lowpass at ~15khz), wouldn't 192 still be a generous cap at V5?

You can also try encoding it in MPC since it's bitrate is somewhat dependant on lowpass filtering and stuff… If -q5 gives you a good result, then just rename it to *.mp3.mpc and that's all.

I wouldn't rely too much on auCDtect...




and check this out: http://www.hydrogenaudio.org/forums/index....ndpost&p=286836

I've got an interesting information.

I had an mp3 (3.90.3 APS ripped) allowing freq. up to 19.5 KHz
It was 207kbps

Then I applied LPF of 16KHz and burnt a AudioCD using that MP3. (Just like the case of original poster)

Then I ripped it again using 3.90.3 APS only. And what I got was VBR mp3 (of cource), with 16KHz cut-off (but obviously). And bitrates were 135kbps.

So I think APS algorithm decided to encode this signal at lower kbps.

Hence, There is no problem using APS. It correctly identifies the need and encodes like-wise.

FYKI:
Song: "Rock The Party" by Bombay Rockers

and



Regards,

M

I'm no expert by any means, but wouldn't alt-preset standard give you the best quality in every scenario? Doesn't re-encoding lower the bitrate each successive time, therefore there's no need in setting some lowpass since APS gives you the best sound possible?

Re-encoding can actually raise the bitrate due to added quantization noise during the first encode.

As I explained, I was experimenting! I do not need to set low-pass. I was EXPERIMENTING! I've even concluded that APS selects the best range.

I Think you'd like to check this out.
If you would please download Dirom's Modified Compile (just because it's easy drag and drop. it works same as original one). and drop any APS encoded mp3. let us say test.mp3. the output file will be smaller and with smaller kbps. Do that again and that would be smaller again

original wav - 4.9MB (29 seconds)
test.mp3 206kbps 736KB
test.mp3.mp3 201kbps 722KB
test.mp3.mp3.mp3 199kbps 710KB
test.mp3.mp3.mp3.mp3test.mp3.mp3.mp3.mp3.mp3 193kbps 698KB