Reason 12 - Silence gap at beginning of mp3?

[KGB]

Why does my mp3 export (Export song as audio file) have so much more noticeable latency than my Wav exports (Bounce Mixer Channel and Export song as audio file)? The Wav files sound basically perfect and just as they are in my session. The mp3s are off beat because of the latency.

I just discovered this issue today..

I now have about 100 - 200 useless MP3 files that will be off beat when imported into a client's DAW.

This is kind of a big deal to me.

Can anyone shed light on this?

[StephenHutchinson]

I discovered this ages and ages ago... but in another DAW. When you export to .mp3, you will get a leading "top" (or silence) at the beginning of the file. In my case, I was trying to make seamless loops and this of course messed me up. Exporting them out to .WAV or .AIFF is the only way not to get the "latency". It sucks, but .mp3 is what it is...

[Aquila]

honestly you'd be better off not using a lossy format for futrher DAW work anyway, MP3 should really only be a final product

[KGB]

I discovered this ages and ages ago... but in another DAW. When you export to .mp3, you will get a leading "top" (or silence) at the beginning of the file. In my case, I was trying to make seamless loops and this of course messed me up. Exporting them out to .WAV or .AIFF is the only way not to get the "latency". It sucks, but .mp3 is what it is...

Sheesh. I appreciate the explanation! So I guess I have to Export out as .Wav and then get an mp3 converter. That sucks big ones...smh

[QVprod]

I discovered this ages and ages ago... but in another DAW. When you export to .mp3, you will get a leading "top" (or silence) at the beginning of the file. In my case, I was trying to make seamless loops and this of course messed me up. Exporting them out to .WAV or .AIFF is the only way not to get the "latency". It sucks, but .mp3 is what it is...

Sheesh. I appreciate the explanation! So I guess I have to Export out as .Wav and then get an mp3 converter. That sucks big ones...smh

don’t think that would accomplish anything different. That’s a matter of MP3 in general. But like stated above, If you’re sending files for someone to import into their DAW, you shouldn’t be using MP3 anyway. It degrades the quality. You’d actually want to use Wav.

[StephenHutchinson]

Sheesh. I appreciate the explanation! So I guess I have to Export out as .Wav and then get an mp3 converter. That sucks big ones...smh

don’t think that would accomplish anything different. That’s a matter of MP3 in general. But like stated above, If you’re sending files for someone to import into their DAW, you shouldn’t be using MP3 anyway. It degrades the quality. You’d actually want to use Wav.

QVprod is entirely correct. Converting files to .mp3 will still result in a small gap at the beginning.

[rgdaniel]

I routinely export to both mp3 and wav when I finish a song. I tried my most recent song, dragged both files into reason, they lined up perfectly. Tried the second most recent, and WOAH... huge difference in how they lined up!

So what was the difference? The second most recent song has a tempo automation clip right at the beginning, where it slows down the intro, for one bar, then end of automation clip, and tempo reverts to default (no automation clip).

The thing is, it's the WAV file that plays incorrectly, sounds like it gets stuck on that temporary slow tempo for the whole song. When you play them back OUTSIDE Reason, they are both identical, but when you drag them INTO reason, the WAV file is a good thirty seconds longer, since it's playing at the slower tempo that was only supposed to last one bar...

I suspected that if I changed the tempo automation clips to always be there with some value, instead of letting it default, that would line them up again, but scratch that theory, tried a few things, including no tempo automation, and the wav file still plays slower than the mp3. Apparently regardless of tempo automation clips after all... Confusing...

Again, both MP3 and WAV are the same when played outside Reason, eg. Windows Media Player, but when you drag them into Reason, something goes wrong with the WAV file, and it plays more slowly. (in my case at least).

Not seeing any leading silence as StephenHutchinson describes.

[rgdaniel]

Tried my third-most-recent song, dragged in the mp3 and the wav into a new reason file, the wav clip is longer and plays slower (incorrectly so). The mp3 plays as it should. Both sound the same and correct when played in Windows Media Player. So tempo automation used in this one.

Tried my 4-recent song, same deal again, WAV file is noticeably sluggish when dragged into Reason. MP3 is fine. Both are fine outside Reason.

And here's the real puzzler. I tried a song from 2017, SAME THING HAPPENED. Wav file is too slow when dragged into Reason. Tried various pairs created in R12, and going back to R10. A few were identical, as you would expect, but many (most of my small sample of a dozen or so) showed the same thing where the WAV played back sluggishly. A difference of maybe 30 bpm if I had to guess. All played fine outside reason.

More info: I opened one of the sluggish WAV's in Audacity, exported immediately to a new WAV on the desktop, dragged THAT into Reason, and it played normally. Weird.

[rgdaniel]

And finally (because I give up) I tried this on Reason 11 on my other PC, and same thing. Files even as far back as 2017. Same results as on my main PC, WAVs were sluggish when dragged into reason compared to mp3's. So not a R12 thing if the problem occurs when wavs exported from R10 and dragged into R11 show the issue. Not ALL files, but many.

[Pepin]

Afaik gapless mp3 is possible if you use an encoder that stores the padding as metadata and a decoder that respects that metadata.
But there's not one standard solution between Apple's encoder, LAME, etc.

[selig]

And finally (because I give up) I tried this on Reason 11 on my other PC, and same thing. Files even as far back as 2017. Same results as on my main PC, WAVs were sluggish when dragged into reason compared to mp3's. So not a R12 thing if the problem occurs when wavs exported from R10 and dragged into R11 show the issue. Not ALL files, but many.

Long shot:
Reasons WAV files have tempo data - try disabling stretch after importing and see if that makes any difference.

[PhillipOrdonez]

That sounds weird, rgdaniel. Have you tried turning off the stretch algorithm?

Must be something in the metadata that tells Reason to do that, since it sounds like it happens only with tempo automated projects. Using another editor for bouncing again probably erases that

[PhillipOrdonez]

Aaand Selig beat me to it.

[rgdaniel]

Long shot:
Reasons WAV files have tempo data - try disabling stretch after importing and see if that makes any difference.

And just like that, my sanity is restored... I had a feeling about stretch, but couldn't see how it would help AFTER the fact of importing it, so I didn't even try it... well I tried it now, and the waveform jumped back to the right length... clip boundary itself still extends beyond that, now empty at the end bit, but tempo is now audibly correct and the waveform lines up with the mp3 version.... Thank you Giles!

This begs the question, do people routinely "disable stretch" when dragging in WAV files? Doesn't come up much for me, I don't work with stems, the closest I come is dragging in files from Loopcloud, then drag-snapping those clips to the correct number of bars. I don't "disable stretch", as such.

[KGB]

Sheesh. I appreciate the explanation! So I guess I have to Export out as .Wav and then get an mp3 converter. That sucks big ones...smh

don’t think that would accomplish anything different. That’s a matter of MP3 in general. But like stated above, If you’re sending files for someone to import into their DAW, you shouldn’t be using MP3 anyway. It degrades the quality. You’d actually want to use Wav.

I know but In my genre no one really wants to deal with wavs, especially through email. MP3 is more convenient and pretty much the standard.

[RobC]

Yep, the beginning silence is an mp3 thing. I used to know why, but forgot. : P

[Tinnitus]

Sheesh. I appreciate the explanation! So I guess I have to Export out as .Wav and then get an mp3 converter. That sucks big ones...smh

don’t think that would accomplish anything different. That’s a matter of MP3 in general. But like stated above, If you’re sending files for someone to import into their DAW, you shouldn’t be using MP3 anyway. It degrades the quality. You’d actually want to use Wav.

Agreed, but FLAC would of course be the best lossless format for storage and export import but sadly Reason still doesn't support this in Reason 12.

[PhillipOrdonez]

don’t think that would accomplish anything different. That’s a matter of MP3 in general. But like stated above, If you’re sending files for someone to import into their DAW, you shouldn’t be using MP3 anyway. It degrades the quality. You’d actually want to use Wav.

I know but In my genre no one really wants to deal with wavs, especially through email. MP3 is more convenient and pretty much the standard.

The standard for consumers*

Has your genre not heard of WeTransfer? They should check it out

[Dabbler]

Wish the title of this thread was a little bit more specific

[rgdaniel]

That sounds weird, rgdaniel. Have you tried turning off the stretch algorithm?

Just noticed that you also nailed this, so thanks for that!

[joeyluck]

Wish the title of this thread was a little bit more specific

Wish granted

[KGB]

I exported as a wav and used Audacity to convert to mp3 and it’s perfect. Why doesn’t the Reason mp3 export trim the silence gap at the beginning of the mp3?

[jonnyretina]

I exported as a wav and used Audacity to convert to mp3 and it’s perfect. Why doesn’t the Reason mp3 export trim the silence gap at the beginning of the mp3?

You must be mistaken. MP3s by the very nature of the file format will always leave a small silent gap at the beginning and the end of the file. This is because the MP3 compression algorithm itself leaves that space of 10ms to 50ms there - regardless of the software doing the compression.

The only way around this is to use FLAC instead of MP3. Alternatively I think OGG might support a lossy compression where it doesn't add small padding around the file but I'm not sure as I don't really use OGG. FLAC FTW.

[KGB]

I exported as a wav and used Audacity to convert to mp3 and it’s perfect. Why doesn’t the Reason mp3 export trim the silence gap at the beginning of the mp3?

You must be mistaken. MP3s by the very nature of the file format will always leave a small silent gap at the beginning and the end of the file. This is because the MP3 compression algorithm itself leaves that space of 10ms to 50ms there - regardless of the software doing the compression.

The only way around this is to use FLAC instead of MP3. Alternatively I think OGG might support a lossy compression where it doesn't add small padding around the file but I'm not sure as I don't really use OGG. FLAC FTW.

You're right but Audacity trims the silence in the beginning of the mp3 when generating the file.

[jonnyretina]

You must be mistaken. MP3s by the very nature of the file format will always leave a small silent gap at the beginning and the end of the file. This is because the MP3 compression algorithm itself leaves that space of 10ms to 50ms there - regardless of the software doing the compression.

The only way around this is to use FLAC instead of MP3. Alternatively I think OGG might support a lossy compression where it doesn't add small padding around the file but I'm not sure as I don't really use OGG. FLAC FTW.

You're right but Audacity trims the silence in the beginning of the mp3 when generating the file.

AFAIK there's no way it could do that, the spacing is always going to be inherent in the file format itself. If you're finding when importing MP3 into Audacity that there is no space, I imagine it is trimming it on import in the newer versions. It certainly didn't used to do that, see this forum post for example: https://forum.audacityteam.org/viewtopic.php?t=95625

A technical explanation about the added MP3 milliseconds of silence can be found in this LAME MP3 FAQ https://lame.sourceforge.io/tech-FAQ.txt

The only way I can imagine this is fixed on export is if LAME have now solved this issue of milliseconds of silence added and Audacity now comes with such a version of LAME. If that is the case, then Reason Studios need to update the version of LAME that Reason is using.