TRUE BITRATE - how to calculate bitrate of mp3, page 2

bottom left you can see the spectra from the richie hawtin set from sonar in the recent megapack ive upped

the middle one is the 128k version from the tracker single file

the right one is a 320k mixing.dj version

any comments??!!

thanks for the first proper report on using the tool!! very nice comparative with pics!

as usual, mixing.dj sucks ass.. =)

that's eye opening, if the 128K version is essentially the same as the 320K version, as you prove on this.  Its a waste of space.

yup

exactly the same spektra and waste of space

and time

..and money?!

technically the same quality

I love how people put EM's in 320 when 99% (cept the original source file provided by the artist, usually on soundcloud) when the Radio 1 Digital stream max is 192

thats right!

the fm is something diferent tho, not digital stream

Whilst we are on the comparison of mixes via spectrum.... Does anyone know how to find out the quality of a stream, especially those video, mixcloud and soundcloud that are floating about at the moment?

Cheers Bidona, great presentation, more ammunition against mixing.dj bollocks....

Besides soundcloud and my own recordings there is not much else i want these days. The quality and the quanity that we got here is just about perfect!

 

soundcloud stream is always 128kbps. mixes are downloadable from there, they are 128kbps no matter what the source had been.

as for mixcloud, they say, it's also extractable somehow, i am not sure exactly. but to check the rate, simply record a piece of sound while it plays on your speakers, and then analyze it. otherwise Tokes should know how to download the stream there. if you do decide to check mixcloud this way, i would recommend trying on several different mixes, better yet - register there and upload 3 tracks of 128, 192 and 320kbps each, and then play/record/download them and analyze, i am ready to bet they'll all be 128kbps or less. also mades knows about it, i think he mentioned it somewhere.

Bidona proved it for me....

This is a little of track but still related to the original thread....

Has anyone trialed/used any codec outside of mp3 for their music files? Particularly i am interested in the comparison of sound quality between MP3 and AAC, particularly in lower quality streams (128k).

I know SpasV used it sometime ago with success and used to swear by its quality.

I will have the true results though some spectrum tests next week, but wanting to hear what other users think. Slash are we able to upload AAC format?

obviously, we're able to upload anything.

I think it's down to encoders rather than strictly the codec itself, at least until someone makes an AAC encoder to beat LAME.

Assuming you are making an MP3 using LAME (3.9X) with optimal settings (very important) the general consensus seems to be that it is basically ~ on a par with FHG/iTunes/Nero's AAC/MP4 encoders.

If you are using something else to encode your MP3s (Blade/FHG/QDesign/X) - then it is likely to be not quite as good or rather shite.

If you are interested in this sort of thing, a really good forum is Hydrogen Audio : -

https://www.hydrogenaudio.org/forums/

They have hundreds of threads discussing this and similar topics. Such as MP2 vs MP3 etc.

It's where I have generally gone to form all the opinions that I've listed above. (Plus using my ears!)

The best thing about the site is that they insist on people making double blind ABX tests on a music sample before making claims (and using the same equipment for the tests).

As a result, most of the claims/conclusions on there are not due to 'placebo' effects.

E.g. 'I made a 320K rip of a CD using LAME/MP3 and it sounds better than/worse than/same as 320K AAC' .... here are the ABX test results to prove it etc etc.

So the site is generally bull-shit free.

After reading through the forums I did some of my own ABX testing and found that at about CBR 128K and above, LAME MP3 is as good as an FHG/iTunes AAC encoder............ I also discovered that the much maligned MP2 codec (using latest TwoLame encoder) sounds as good as an MP3 once you get to CBR 192K and above, for most EDM I tested it with.

But it's not the case with classical music, which seems to catch MP2 out a bit on the higher frequencies, until you go to say CBR 256K, at which point I find it doesn't matter if you are using MP2, MP3 or AAC. They all sound the same to my humble ears.

Some of my friends claim they can tell the difference between an MP3 encoded at CBR 256K v 320K, but having done tests myself, my ears cannot tell the difference between CBR 192 vs 256 vs 320K!!!!! Maybe my ears are fucked, I dunno?!

Another point, something I read on Hydrogen Audio, is that if you re-encode say an MP3 at 128K to 320K (coughs) like Mixing.DJ then there will be a slight quality DECREASE in the newly encoded 320K MP3 vs the original 128K MP3, This is because of the fact that MP3 is lossy and uses cosine transforms. Whatever the heck they are.

So not only are these direct download sites ripping you off by giving you a bigger (more expensive) file to download, it is also slightly less quality than the 128K version it was encoded from. The  difference is very very slight, but it's there nonetheless......... booooooooh! Hisssss!

Cheers Rich, I had been over to Hydrogenaudio forums, read some of the posts on it and seen some of their comparisons, makes for some interesting consideration.

Like you,  most of the time my ears can tell the story of quality. My main ponder is that i could offer the 128k AAC frisky stream rather than the 128k mp3. At this stage my ears can't tell the difference in between the two. I will continue this week to put two files on spectrum analysis and see what i come up with.

That is why we are here and not over there, yes it is a very greedy decision to up-encode whilst also being bandwidth hungry. I bet the ISP companies with caps etc love sites like mixing.dj up-encoding their files. Interesting point that up-encoding reduces audio quality even more through the conversion process, bet you that hasn't been taken into consideration.

thanks, RichTee!! good info!!

i liked today's example of upencoding.. several acts from Rock In Rio Festival were broadcast on youtube. and recorded.. and ended up being shared all over the net by direct download sites as 320kbps files. while youtube is a video streaming platform that doesn't really care about audio, thus the latter streams at most 96kbps.. 320 vs 96.. of my fcking god...

and another example, one of those sets i came across was a scene release.. group EITHELMp3 has released 2 versions of Carl Cox @ Rock in Rio 2012, both of them were VBR average at 224, both of them had SAT aka Satellite as their source... another OMFG.. they gotta be kidding? is it an error? or are they so arrogant to put source SAT, while it's a simple stream.