Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Mp3 artifact questions and comparison (Read 10673 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Mp3 artifact questions and comparison

I am trying to determine if some music I have from artists are bad quality recordings or bad bitrate but I had some general questions as well

This is taken from the mp3 artifcats wiki

"When the psychoacoustic model fails, when the transform block size is restrained, or when aggressive compression is used, this may result in compression artifacts. Compression artifacts in compressed audio typically show up as ringing, pre-echo, drop-outs, warbling, metallic ringing, an underwater feeling, hissing, or "graininess".

A good way to observe compression artifacts in audio is to listen to the applause in a relatively highly compressed (eg 96 kbit/sec) MP3. In general, music has repeating structures and more predictable variations in volume, whereas applause is essentially random, therefore hard to compress. This highly compressed track of applause will illustrate the "metallic ringing" and other compression artifacts very well."

Now are there artifacts in all mp3s just at a lesser rate in higher quality mp3s?

Musically is there anything different between a 128 kbps file VS 192 KBPS or just the amount of artifacts?  Or is there less detail in 128?

I think I know what ringing is, what exactly is a pre-echo? 

Is a drop out when the music just stops?

Is graininess like distortion?

Metallic ringing I am assuming is bad cymbals, etc?

What exactly is warbling?


I use a pair of $500 Shure Headphones for my IPOD will that make me hear artifacts easier, if so why?


Thanks experts!

Mp3 artifact questions and comparison

Reply #1 –
Let's start with the definition of artifact.
An artifact is an unwanted, unpleasant effect of encoding, which happen mostly due to two main causes:
a) Not enough bitrate: The encoder cannot do its best, so it generates a distortion.
b) Erroneous analysis (Bug or inadequate parameters): The encoder thinks it's good to add the distortion.

Generally speaking, with a higher bitrate, there's less probabilities of "a", but "b" can still appear.
(This sort of explains your question between 128 and 192kbps. Less bitrate is more prone to artifacts and obviously has less audible information).

Now, onto types of artifacts:
First, you may want to visit this page: http://www.ff123.net/training/training.html . It isn't updated (so it isn't using current encoders), but just like simple maths are always the same, artifacts are always artifacts.

Next, pre-echo is hearing something before it actually starts. When codecs don't use the appropiate block size on drums, hihats, claps, fast paced notes or other similar scenarions, the beginning of the sound is no longer precise, but distorted, noisy, sometimes like a volume slide. Castanets on that training page is a pre-echo example.

Metallic ringing not only appears with cymbals. WMA (standard) at 64kbps (stereo) is a good example of a codec that produces Metallic ringing on many samples.
Warbling is an effect that perceively sounds like underwater sound. It is similar to a drop out, but warbling is more repetitive.


About headphones, they tend to help on hearing artifacts, especially if they are closed, reducing the noise from outside.

[Edit: explained it a bit more]

Mp3 artifact questions and comparison

Reply #2 –
Quote from: airtas on
I think I know what ringing is, what exactly is a pre-echo?


Pre-echo is a artifact that appears due to wrong block usage or using very low bitrates. It can produce a "psst" noise and produces smearing on sharp transients samples, such as castanets.

Pre-echo will likely to bother you, if your a big fan of drums, castanets, live music (auidence clapping) and electronic music.

A few precho samples that are very easy to ABX at 320kbps due to pre-echo:

EIG
Show Me Your Spine
Human Disease (ABXable at LAME -V2)

Quote from: airtas on
What exactly is warbling?


Warbling is the most common form of artifacts IMO. It produces metallic watery noises, making a track sounding like a "Splish Splash Show".
"I never thought I'd see this much candy in one mission!"

Mp3 artifact questions and comparison

Reply #3 –
Quote from: /mnt on
Quote from: airtas on
I think I know what ringing is, what exactly is a pre-echo?


Pre-echo is a artifact that appears due to wrong block usage or using very low bitrates. It can produce a "psst" noise and produces smearing on sharp transients samples, such as castanets.

Pre-echo will likely to bother you, if your a big fan of drums, castanets, live music (auidence clapping) and electronic music.

A few precho samples that are very easy to ABX at 320kbps due to pre-echo:

EIG
Show Me Your Spine
Human Disease (ABXable at LAME -V2)

Quote from: airtas on
What exactly is warbling?


Warbling is the most common form of artifacts IMO. It produces metallic watery noises, making a track sounding like a "Splish Splash Show".




So artifacts can sound like "clipping/distortion" in a way?  On some lower bitrate MP3's I feel like all the sound is mushed together, like there is no channel or instrument seperation.

Does that make sense?

Mp3 artifact questions and comparison

Reply #4 –
Quote from: airtas on
On some lower bitrate MP3's I feel like all the sound is mushed together


Sounds like smearing due to pre-echo artifacts. You will need to do a ABX test to make sure that it wasn't a placebo though.
"I never thought I'd see this much candy in one mission!"

Mp3 artifact questions and comparison

Reply #5 –
Quote from: airtas on
Now are there artifacts in all mp3s just at a lesser rate in higher quality mp3s?


Good question. I guess since all mp3s have blocking issues to a certain degree, they all also should have a certain level of artifacts. But the real question is; is it possible for the normal human ear to spot artifacts in all mp3 samples? And this question, I would answer with a clear "No". There are a lot of things which our eye's can't see and our ears can't hear!


Quote from: airtas on
Musically is there anything different between a 128 kbps file VS 192 KBPS or just the amount of artifacts?  Or is there less detail in 128?



Well, at least when using mp3, most people with semi-decent ears and equipment should be able to hear the difference in most samples. It is not just the probability to come across artifacts when listening, such low bitrate mp3 samples lack dynamic and sound similar as listening to music through very very poor equipment even when using decent stuff. The less decent the sound processing equipment (and I count the ears in as part of the signal chain), the less likely is it to hear a difference.



Quote from: airtas on
I use a pair of $500 Shure Headphones for my IPOD will that make me hear artifacts easier, if so why?


As far as ears and player aren't the bottleneck here, better headphones will increase the chance to spot artifacts by offering a more detailed listening experience. For what are artifacts but small details? 


MFG

Mp3 artifact questions and comparison

Reply #6 –
Are highs only removed or are bass, and mids also compressed?

Mp3 artifact questions and comparison

Reply #7 –
Quote
I am trying to determine if some music I have from artists are bad quality recordings or bad bitrate...
Without an uncompressed reference, that can be difficult.  You can have something that sounds like a compression artifact, but is actually caused by something else.

Quote
Are highs only removed or are bass, and mids also compressed?
Everything is compressed.  As I understand it, lower frequencies are "easier" to compress mainly because they are less "transient" and less "complex".    At any reasonable bitrate, you should be able to perfectly compress and then reproduce a 15kHz constant tone.  (I haven't actually tried this.)  But, when you start adding dynamics and harmonics (EDIT - or overtones & sub harmonics...  You don't have to worry about harmonics at 15kHz!), the waveform contains more "information" and compression gets trickier.

The highest frequencies are sometimes filtered-out first (depending on the encoder options/settings) because you can't hear them anyway.  This allows the compression algorithm to focus on things that you can hear.