Export (multi) with too big difference in sound .need HELP

Hello, I cut an mp3 file (44k, 24bit) I create my markers and when I export the different markers, I have quite strong differences in sound. if I have an acapella, or with a very low background sound, the sound is increased, if I have strong enough lyrics and bass, the sound is crushed …
I specify that I already adjust the volume before not saturate. I tried using export MP3 or ogg (I need ogg) and the result is the same, I had a less big difference when switching to 24bit (then not floating) but it still gives me a audible difference that prevents me from using my ogg as I need it.
Thanks in advance

I am using the latest version of Audacity (as of February 1, 2021), I have windows10,

(It’s translate text, i’m sorry if it’s really bad english)

PS, when i use all little OGG created, and past all for create new song, i listen all difference too in audacity.

I cut an mp3 file (44k, 24bit)

24kbps bit rate? (MP3 does not have a bit depth like a 16 or 24-bit WAV file). 24kbps would be “low quality” and it’s not good for music.*

When you open a compressed file in Audacity or any “regular” audio editor it gets decompressed. If you re-export as MP3 you are going through another generation of lossy compression and “damage” does accumulate. AAC (MP4/M4A) is immune to accumulated damage, but that’s only true if you start with an AAC file. I don’t know about OGG, but OGG is also lossy and the MP3-to-OGG conversion is a lossy-to-lossy conversion and it’s going to accumulate damage.

Information is thrown-away during compression to make the file smaller. A lower bitrate gives you a smaller file (more compression) and generally lower quality. No damage is done during _de_compression so if you export as WAV or another lossless format there will be no further deterioration. (It gets decompressed when you play it anyway.)

Ideally, you should work in a non-lossy format and compress ONCE as the last step, if you want a lossy format.

If you’re stuck with a lossy original you can minimize the damage by exporting to a higher bitrate (or “higher quality” setting) and you should ALWAYS try to minimize the number of generations… If you want to change something go-back to the original. And, save any intermediate-temporary files in a lossless format.

There are special-purpose MP3 editors (such as MP3directCut) that do not decompress the data. But, that’s not going to help if you want OGG.



\

  • Just as a “reference” - I think MP3s from Amazon or AAC files from iTunes are 256kbps. The maximum for MP3 is 360kbps. We usually don’t “talk about” the bitrate for uncompressed audio, but CDs are 16-bit, 44.1kHz, 2-channel stereo and from that we can calculate 16 x 44.1k x 2 = 1411kbps. With FLAC (lossless compression) we can get down to about half the original bit rate, of course with no loss of quality.

Hello Doug, Ty for reply
i’m sorry, i used bad word…
i’m in stereo, 48000Hz and 24 bit PCM

i know OGG or MP3 is no best quality… my problem is :
if i export my marker in multi (1.ogg, 2.ogg, 3.ogg (level 5 or 10(max)) or .MP3 (best quality) sound change “volume” it’s compressed or if only voice it’s verry hight.
if i export in one file, MP3, ogg or wav… No lost qulity, no compressed or hifght sound.
You understand ?

i create one share folder with screen of audacity and 2 MP3.

BeforeExport.mp3 is one short version of original before i export to multi,

TestPlayMarker.mp3 is when i play my “export file” in Virtual DJ and record ( i cut part of song too)

https://app.box.com/s/6bz11s0kq2ubqpw60ea0ix13c1zvbasf

i hope you can help me with this exemple, if need i can share too the “multi export” it’s +/- 15sec marker

OK. So it looks as though you are creating labels via Tools > Regular Interval Labels ever 15.0 seconds.

I have downloaded your BeforeExport.mp3, created labels then did an Export Multiple of several .ogg files. I then did an import multiple in Audacity then did an Tracks > Align Tracks > Align End to End, followed by a Select All then Tracks > Mix > Mix and Render. I see no visible difference whatsoever between the exported and imported tracks.

So you, instead, are playing the audio via Virtual DJ and re-recording the output via Audacity. Is that correct ?

Is it possible that either Windows or Virtual DJ are somehow trying to normalize or enhance the sounds ?

For Windows. Right-click on the speaker in the Windows Task Bar near the lower-right-hand corner of the screen. Select Sounds > Recording > [Your Playback Device] > Properties > Enhancements. Check that all enhancements and boosts are turned off. Repeat for Playback.

For Virtual DJ, right now you are on your own.

I hope this helps. :smiley:

i record in virtual dj, but i just record for show difference. i cut 15 sec when is 128 bpm, it’s on “beat” …
i use that for create in live “remix” of music, and i use that in one game online too, need only use ogg on this game.
but i change sound card mapper, and now it’s not so bad. ty for help Jademan :slight_smile:

You are welcome. :smiley: