Hello.
Today I noticed that Vegas unnecessarily cuts the frequency band when rendering from Flac to mp3 320kb.
Tested on two songs.
The first image is the original Flac.
The second image is Flac converted to mp3 320kb in Aimp3 player.
And the third image is Flac to mp3 320kb in Vegas.
It is normal?
Or does Vegas use an outdated mp3 plugin?
Thanks for your feedback.
Oh, I see your player is using CBR. You may want to compare file sizes.
Posting your MediaInfo properties will be a big help. I did all of my conversions with LAME in CDex
Oh, I see your player is using CBR. You may want to compare file sizes.
Posting your MediaInfo properties will be a big help. I did all of my conversions with LAME in CDex
Both files are the same size, but the Vegas frequency is cropped.
Vegas uses the Fraunhofer codec.
According to the charts, it's more like between 192kbs and 224kbs.
AIMP3 320kbs:
General
Lennox - Little Bird (CD Single).mp3
Format : MPEG Audio
File size : 10.6 MiB
Duration : 4 min 38 s
Overall bit rate mode : Constant
Overall bit rate : 320 kb/s
Album : Little Bird (CD Single)
Album/Performer : Annie Lennox
Track name : Little Bird (Single Version)
Track name/Position : 01
Performer : Annie Lennox
Genre : Pop
Recorded date : 1993
Writing application : ExactAudioCopy v1.1
Writing library : LAME3.100
Cover : Yes
Cover type : Cover (front)
Cover MIME : image/jpg
Audio
Format : MPEG Audio
Format version : Version 1
Format profile : Layer 3
Duration : 4 min 38 s
Bit rate mode : Constant
Bit rate : 320 kb/s
Channel(s) : 2 channels
Sampling rate : 44.1 kHz
Frame rate : 38.281 FPS (1152 SPF)
Compression mode : Lossy
Stream size : 10.6 MiB (100%)
Writing library : LAME3.100
VEGAS 320kbs:
General
Format : MPEG Audio
File size : 10.6 MiB
Duration : 4 min 38 s
Overall bit rate mode : Constant
Overall bit rate : 320 kb/s
Genre : Other
Audio
Format : MPEG Audio
Format version : Version 1
Format profile : Layer 3
Duration : 4 min 38 s
Bit rate mode : Constant
Bit rate : 320 kb/s
Channel(s) : 2 channels
Sampling rate : 44.1 kHz
Frame rate : 38.281 FPS (1152 SPF)
Compression mode : Lossy
Stream size : 10.6 MiB (100%)
Interesting finding. Which tool do you use for the spectrum analysis?
Interesting finding. Which tool do you use for the spectrum analysis?
Spek
I've edited a few Flac songs in Vegas before and rendered to mp3 320kbs.
I believed that 320kbs is a guarantee of maximum mp3 quality.
Now, out of curiosity, I put the files into the analyzer and I found this weird result.
Writing library : LAME3.100
As I hinted before, I use LAME for all of my mp3 conversions. Suggest you do the same.
Writing library : LAME3.100
As I hinted before, I use LAME for all of my mp3 conversions. Suggest you do the same.
There are many such similar proposals.
Use Vokouder or Happy Otter scripts to render the video. Is better.
And now use Lame codec to render mp3. Is better.
What's next?
Isn't that weird with professional software?
No it's not weird, as some of the best codecs are open-source. But Vegas has licen$ing agreements going back a quarter-century, including Fraunhofer.
I just tested on my own and used various HD audio 24/96 FLAC input files. Encoded to MP3 via AIMP (LAME) and Vegas Pro.
I could not detect any spectrum differences which would lead to a quality drop in the VP version. The differences are in an area above 16 kHz und below – 60 dB. I think the Lame encoder actually wastes bandwidth there as a human can't hear such sound.
The frequency axis goes up to 23 kHz in this window(s), the level axis down to – 66 dB.
Edit:
I accidentally used 48 kHz for the AIMP LAME version and 44,1 kHz for the Vegas Pro version. Maybe I'll demo an updated version later, though there won't be any further disadvantages of the VP version.
See this for instructions to confirm the Fraunhofer encoder is actually' enabled.
As MV suggested, I also prefer the 'LAME' encoder. I like the WinLAME rc3 UI. which supports mono encoding up to 320kps. By changing 'presets' file, many other parameters can be adjusted, HP/LP filters, ect., ect' which are not available on most MP3 encoders.
Otherwise, the MP3 format inherently has a max bitrate of 320kbs... however that can be 'cheated' if spacial content is not present or needed (spoken word for instance), by encoding in a mono MP3 mode. example, a 192kbs file would have the equivalent quality of a stereo 320kbs, or... a mono 320kbs would have the theoretical resolution of a 740bps stereo MP3. In my sum/difference spoken word file tests, a 48/16 PCM and 320kbs CBR MP3 totally cancel out, which indicates little to no audible quality difference in a spoken word mono files.
I just tested on my own and used various HD audio 24/96 FLAC input files. Encoded to MP3 via AIMP (LAME) and Vegas Pro.
I could not detect any spectrum differences which would lead to a quality drop in the VP version. The differences are in an area above 16 kHz und below – 60 dB. I think the Lame encoder actually wastes bandwidth there as a human can't hear such sound.
In Vegas, MP3 320kbs will cut me all.
Wav. Flac. 48khz 44khz. It does not matter.
Marco, what version of Vegas do you have?
Try using my analyzer.
Is there any chance of enabling the Lame codec for rendering in Vegas?
Possible question for the Vegas team for the future version:
Can the user have a choice of Lame or Fraunhofer when rendering?
I used Vegas Pro 17 and I used various different analyzers which all states the observation. Even yours (Spek) does. Forget about everything blue in Spek, it's – 80, – 90 dB and much lower. You won't hear such silent tones no matter which frequency. And forget about the greenish parts (which then is – 60 dB and lower) if it's above 16 kHz. You won't hear that.
Spek is quite misleading for a quality analyze of MP3 audio because there's no way to filter parts which are just useless for a distribution format like MP3. I showed the SIR Spectrum Analyzer in that video above on very purpose.
By the way – I think the Vegas Pro MP3 encoder is not even the Fraunhofer one but an own one (from the days of Sonic Foundry or Sony). And in my opinion it's not a bad choice for the MP3 encoder to let it cut-off low powered 16+ kHz except for certain peaks.
I used to have the LAME encoder available in Sound Forge, but it was years ago and I do not recall the set-up. I use the Fraunhofer encoder for low-res files, since they all sound pretty awful below 128kbs (stereo). Otherwise the WinLAME UI which does batch work as well.