How to get best result when creating the MP3?
-
- KVRAF
- 6207 posts since 25 Dec, 2004
i actually not arguing, I was just saying it's a pointless argument
sketches... http://soundcloud.com/onesnzeros
some artists i support... https://bandcamp.com/spectraselecta
some artists i support... https://bandcamp.com/spectraselecta
-
- KVRAF
- 5646 posts since 18 Jul, 2002
Yes, just bounce or save as MP3 using your host or audio editor at 320 kbps and you're pretty much set.
-
- KVRAF
- Topic Starter
- 5440 posts since 4 Aug, 2006 from Helsinki
If you say it's pointless, you should give more backgroung information,sqigls wrote:i actually not arguing, I was just saying it's a pointless argument
technical arguments etc. Anyone can claim anything but it doesn't
bring us any further in the discurse if you don't have any knowledge behind you statement.
If you just say "I can't hear the difference between the dithered and non-dithered file", OK, it's good to know that you have this experience but it doesn't prove that dithering is not useful.
Your single experience don't explain the phenomenon/the theory if you can't give more general explanation and show that you have the technical know-how
which support your argumentation.
On my behalf I say that I don't know if the dithering is useful or not. My point in this thread is not the usefulnes of dithering - the point is: 1) more information how you generate your MP3 from your DAW and 2) IF you use dithering in CD-formate, should we also use it in MP3 - why and/or why not? H.
-
- KVRAF
- 13964 posts since 19 Jun, 2008 from Seattle
+1Mushy Mushy wrote:I click "Save as mp3" in SoundForge. Job done.Harry_HH wrote:How do you create your MP3 of your own productions?
I'm not a musician, but I've designed sounds that others use to make music. http://soundcloud.com/obsidiananvil
-
- KVRAF
- Topic Starter
- 5440 posts since 4 Aug, 2006 from Helsinki
I.e. your point is that everyone should use the SoundForge? I don't believe that this helps all those Cubase, Logig or Live users very much.Shabdahbriah wrote:+1Mushy Mushy wrote:I click "Save as mp3" in SoundForge. Job done.Harry_HH wrote:How do you create your MP3 of your own productions?
And please tell us who do not use SoundForge when you "click as MP3", is there any option for MP3 dithering - and if you have ever though why/why not?
All this kind information is more useful than +1. H.
-
- KVRAF
- 16861 posts since 8 Mar, 2005 from Utrecht, Holland
+1BertKoor wrote:I think it's possible to make MP3 files directly from 24bit or 32bit masters.
That skips the dithering stage.
Me thinks it's all irrelevant anyway. Tracks are so damned loud today that you won't hear a difference with truncated instead of dithered versions. Add to that that when you listen to MP3 files, you're hearing a reconstruction anyway, only resembling the original. Minute details like dithering noise fall away pretty quickly.
We are the KVR collective. Resistance is futile. You will be assimilated. 
My MusicCalc is served over https!!
My MusicCalc is served over https!!
-
- KVRAF
- 15135 posts since 7 Sep, 2008
YesHarry_HH wrote:I.e. your point is that everyone should use the SoundForge?
And I use Cubase & Live from which I export at 24/48 into SoundForge.
No. Don't care. If it needed it SoundForge would do it.Harry_HH wrote:And please tell us who do not use SoundForge when you "click as MP3", is there any option for MP3 dithering - and if you have ever though why/why not?
"I was wondering if you'd like to try Magic Mushrooms"
"Oooh I dont know. Sounds a bit scary"
"It's not scary. You just lose a sense of who you are and all that sh!t"
"Oooh I dont know. Sounds a bit scary"
"It's not scary. You just lose a sense of who you are and all that sh!t"
-
- Banned
- 395 posts since 21 Dec, 2010 from Vermont, USA
sure...Harry_HH wrote:...Can you kindly elaborate...
-
- KVRAF
- 13964 posts since 19 Jun, 2008 from Seattle
Yes. Or WaveLab, or Audition. If those apps are not available/do-able, then (personally) Wavosaur over Audacity.Harry_HH wrote:I.e. your point is that everyone should use the SoundForge?Shabdahbriah wrote:+1Mushy Mushy wrote:I click "Save as mp3" in SoundForge. Job done.Harry_HH wrote:How do you create your MP3 of your own productions?
Bummer.Harry_HH wrote:I don't believe that this helps all those Cubase, Logi[c] or Live users very much.
No. There is no such "option" - because doing so (as previously, and repeatedly stated) is pointless.Harry_HH wrote:And please tell us who do not use SoundForge when you "click as MP3", is there any option for MP3 dithering
Dude... (may I call you "Dude"? ... thanks)Harry_HH wrote:All this kind [of] information is more useful than +1.
I answered your question on page one.
To recap: You do NOT "dither" TO MP3.
YMMV!!!
I'm not a musician, but I've designed sounds that others use to make music. http://soundcloud.com/obsidiananvil
-
- KVRAF
- 2083 posts since 28 Feb, 2011
A.M. Gold wrote:Huh? In what sense? Do you have to buy an additional Sonar add-on to get it to spit out an mp3?Gonga wrote:This is a good question and I'll follow the responses as I know basically nothing about it! SONAR requires you to pay extra for the MP3 exporting...
Reaper just uses the free LAME encoder.
I have always just used Audacity...no idea about using LAME within SONAR without paying somehow - worth checking into...
-
- KVRAF
- Topic Starter
- 5440 posts since 4 Aug, 2006 from Helsinki
[quote="ShabdahbriahDude... (may I call you "Dude"? ... thanks)
I answered your question on page one.
To recap: You do NOT "dither" TO MP3.
YMMV!!![/quote]
Dude - may I call you Dude? I call you anyway, Dude.
Have you heard about reasoning... or discussion if the word reason
is difficult. Maybe you have accustomed just using words "NOT"
"YIP" locally, but in these forums we are used at lest trying to explain
things - unless you are from Vermont. H.
I answered your question on page one.
To recap: You do NOT "dither" TO MP3.
YMMV!!![/quote]
Dude - may I call you Dude? I call you anyway, Dude.
Have you heard about reasoning... or discussion if the word reason
is difficult. Maybe you have accustomed just using words "NOT"
"YIP" locally, but in these forums we are used at lest trying to explain
things - unless you are from Vermont. H.
-
- KVRAF
- 4801 posts since 1 Aug, 2005 from Warszawa, Poland
I was investigating this issue a while ago, there was my thread somewhere 3 or 4 years ago, anyway, my conclusions were:
MP3 works 32 bit internally (if I remember correctly), so basically it's converted to 32 bit before encoding anyway.
Dithering doesn't matter, it's very subtle compared to distortion/noise introduced by MP3 encoder. Anyway, most encoders will apply lowpass filter by default, unless you turn it off manually, which doesn't really make sense, unless you want 320 kbps maybe, in which case you are better off with loseless compression.
My final point was that encoders are optimized to handle standard CD format, dithered 44.1, 16 bit. This is 99% of input, so better stick to it.
If you have busy "high-end of the spectrum" and you have to go mp3, you could try to tame it manually especially for mp3: less highs = more bandwidth = less distortion. But I don't think it's worth to bother really.
MP3 works 32 bit internally (if I remember correctly), so basically it's converted to 32 bit before encoding anyway.
Dithering doesn't matter, it's very subtle compared to distortion/noise introduced by MP3 encoder. Anyway, most encoders will apply lowpass filter by default, unless you turn it off manually, which doesn't really make sense, unless you want 320 kbps maybe, in which case you are better off with loseless compression.
My final point was that encoders are optimized to handle standard CD format, dithered 44.1, 16 bit. This is 99% of input, so better stick to it.
If you have busy "high-end of the spectrum" and you have to go mp3, you could try to tame it manually especially for mp3: less highs = more bandwidth = less distortion. But I don't think it's worth to bother really.
-
- KVRian
- 659 posts since 25 Sep, 2010
Dither is not applicable to MP3 encoding, period. If you plan to export your 24- or 32-bit project to 16 bits prior to encoding, then yes, apply dither then. But if you're not doing that intermediate step, then dither is not a concern.
Dither compensates for rounding errors when you lose the bottom 8 bits converting from a 24-bit (or 32-bit) wave to a 16-bit wave. This isn't what happens during MP3 encoding. Contrary to popular belief, data is not reduced to 16 bits within the encoder. 24-bit data can be encoded as-is; internally, the data stored in the MP3 file is floating-point. Encoding strips away any data deemed inaudible, and although that does result in wordlength reduction, by definition you cannot hear that loss of data. Any dither that had been present in the source would be discarded along with the other least-significant fluctuations in amplitude that the encoder has decided is unnecessary.
In short, dither is a non-issue when it comes to maximizing MP3 quality.
All MP3 encoders except LAME use the same algorithms, licensed from the Fraunhofer Institute, which owns the patent. LAME independently reverse-engineered the algorithm to legally reproduce functionally-equivalent output without actually using the Fraunhofer code. What this means is that all encoders produce files of comparable quality at a given bitrate.
In my experience, I've heard no difference between files encoded with Fraunhofer-licensed encoders versus LAME. (BTW, that Cakewalk encoder they charge extra for is actually LAME in disguise.)
Adobe Audition is slightly different, in that it supports an enhanced variant of MP3 called MP3Pro. It stores additional high-frequency information that's normally discarded during encoding, but requires an MP3Pro-enabled player to make use of that data, so most players ignore it. Consequently, AA-encoded files sound no different than those created by other programs (although AA provides more encoder options than many other applications).
So what does have an impact on the quality of an MP3 file? Two things: the bitrate, and the quality of the source data. Just use the highest practical bitrate (although nobody I know can hear a difference between 256kb/s and 320kb/s) for your application.
Even if you're uploading to a site that's going to re-encode your file to a lower bitrate, or streams at a lower bitrate, you'll still want to send them the highest bitrate they'll accept. For example, a free SoundClick account limits files to 10MB and re-encodes them to 128kb/s. Nevertheless, encode your song to the highest bitrate that results in an under-10MB file size.
One last thought. MP3 encoding involves the use of very high-Q filters, which are prone to ringing. This ringing can cause unexpectedly high peak values that could possibly distort a cheap MP3 player. It's therefore a good idea to master your material with more headroom than you might for a CD. I recommend leaving at least 1db headroom when the destination is an MP3 rather than the customary 0.1 to 0.3db or so used with CDs.
Dither compensates for rounding errors when you lose the bottom 8 bits converting from a 24-bit (or 32-bit) wave to a 16-bit wave. This isn't what happens during MP3 encoding. Contrary to popular belief, data is not reduced to 16 bits within the encoder. 24-bit data can be encoded as-is; internally, the data stored in the MP3 file is floating-point. Encoding strips away any data deemed inaudible, and although that does result in wordlength reduction, by definition you cannot hear that loss of data. Any dither that had been present in the source would be discarded along with the other least-significant fluctuations in amplitude that the encoder has decided is unnecessary.
In short, dither is a non-issue when it comes to maximizing MP3 quality.
All MP3 encoders except LAME use the same algorithms, licensed from the Fraunhofer Institute, which owns the patent. LAME independently reverse-engineered the algorithm to legally reproduce functionally-equivalent output without actually using the Fraunhofer code. What this means is that all encoders produce files of comparable quality at a given bitrate.
In my experience, I've heard no difference between files encoded with Fraunhofer-licensed encoders versus LAME. (BTW, that Cakewalk encoder they charge extra for is actually LAME in disguise.)
Adobe Audition is slightly different, in that it supports an enhanced variant of MP3 called MP3Pro. It stores additional high-frequency information that's normally discarded during encoding, but requires an MP3Pro-enabled player to make use of that data, so most players ignore it. Consequently, AA-encoded files sound no different than those created by other programs (although AA provides more encoder options than many other applications).
So what does have an impact on the quality of an MP3 file? Two things: the bitrate, and the quality of the source data. Just use the highest practical bitrate (although nobody I know can hear a difference between 256kb/s and 320kb/s) for your application.
Even if you're uploading to a site that's going to re-encode your file to a lower bitrate, or streams at a lower bitrate, you'll still want to send them the highest bitrate they'll accept. For example, a free SoundClick account limits files to 10MB and re-encodes them to 128kb/s. Nevertheless, encode your song to the highest bitrate that results in an under-10MB file size.
One last thought. MP3 encoding involves the use of very high-Q filters, which are prone to ringing. This ringing can cause unexpectedly high peak values that could possibly distort a cheap MP3 player. It's therefore a good idea to master your material with more headroom than you might for a CD. I recommend leaving at least 1db headroom when the destination is an MP3 rather than the customary 0.1 to 0.3db or so used with CDs.
-
- KVRAF
- Topic Starter
- 5440 posts since 4 Aug, 2006 from Helsinki
To ensure I got this right: you say that best source material for MP3 is dithered CD-format, not e.g. your master file? Can you give me some reference link source of this. H,Zombie Queen wrote:My final point was that encoders are optimized to handle standard CD format, dithered 44.1, 16 bit. This is 99% of input, so better stick to it.