Music Analysis - Objective & Subjective

[The Computer Audiophile]

The Music Analysis forum was created to help Computer Audiophiles make educated purchases of physical formats and downloaded music. This area is for objective and subjective music analysis. Each thread is limited to discussing a single album. Objective analysis can include Spectrograms / FFT graphs, Waveform screenshots, and Dynamic Range information.

 

Thanks to CA reader wgscott [Blog] for starting the several hundred post long thread titled The HD music fft atlas reference thread and for his recent article titled It's Time To Fight Back. The FFT thread is wonderful but becoming a bit too popular for its own good. Browsing the thread can be difficult and inefficient. I created this forum to improve upon wgscotts original idea.

 

In this forum each album will have its own thread and analysis. This style of posting will make browsing and educating oneself immensely easier.

 

In addition to CA readers I highly recommend and encourage Record Labels to publish objective analysis of their music.

 

 

 

Some tools to produce graphs include.

1. Audacity [Link]

2. Adobe Audition [Link]

3. FuzzMeasure Pro [Link]

4. TT Dynamic Range Meter (Standalone) [Link]

5. TT Dynamic Range Meter (Foobar Component) [Link]

[AlainGr]

Hi Chris

 

This is maybe a little early to suggest and I am not sure if it is feasable (and desirable), but...

 

If this is intended to be a sort of "database", maybe to have an hierarchical and alphabetical view of things could help to access the information easier ? I am not sure at what rate the contributions will increase though...

 

Rock -> Pink Floyd -> Wish you were here

Rock -> Hendrix_Jimi -> Experience (I use the underscore when I reverse the name to allow thinks to be alphabetic)

 

?

 

Alain

 

[The Computer Audiophile]

Hi Alain - Thanks much for the input. I'm going to wait for this section to get populated for awhile then make a decision. I will be able to move the threads easily if needed.

 

 

[Sik_Lescinovid]

In any case I'd never try and categorize things by genre since it leaves too much room for confusion. If any categorizing is to be done at any point it should be by artist name and album.

 

Thanks for this new forum Chris, it definitely makes it much easier than in the original thread!

 

[Garf]

I love this idea Chris, much better than loading up the entire fft atlas thread. Even with images cached, it takes a long time to render!

 

Possibly we can add another sticky with guidelines on settings for audacity and other tools as found in the other thread.

 

Similar, guidelines on naming conventions for the post threads can help categorize, for example using genre, album name in the title.

 

[Brian A]

I have been browsing the websites (Audacity etc) that Chris listed in his OP in an earnest effort to teach myself how to interpret the charts that Bill et al have been posting. I am not having much luck.

 

Is there a link to a brief tutorial on how to read the charts (Frequency Spectrum, Frequency Distribution, Waveform etc)?

 

Alternatively, can someone jot down two or three bullet points explaining what each of these charts is telling us regarding the recording? Ideal would be an example chart and then the two or three bullets.

 

This looks like very interesting information.

 

 

[wgscott]

waveform: Allows you to see if the dynamic range has been exceeded.

 

This one looks kind of bad:

 

This one looks good:

 

 

spectrum (fourier transform): Allows you to see how high in frequency the signal actually goes, given the cut-off imposed by the type of file (red-book is sampled at 44.1 kHz, so it cuts off at 22.05 kHz):

 

This one looks bad (probably up-sampled):

 

This one looks good:

 

The third type of plot is basically a cross-section of the second type of plot, but provides a complementary view.

 

 

 

 

[Mike in MD]

another very good analysis tool.

 

http://www.sonicvisualiser.org/

 

Really good for seeing clips.

 

[Jay192]

and special thanks must go to Bill/wgscott who copped some unwarranted flak elsewhere with his Talking Head's analysis.

 

[jcpom]

moved to its own post new post (I hope that is OK.)

 

[Garf]

Bumping this back to the top of the forum.

 

Chris, can we even sticky this thread?

[wgscott]

and special thanks must go to Bill/wgscott who copped some unwarranted flak elsewhere with his Talking Head's analysis.

 

it is ongoing, and directed not only at me, but anyone with the Audacity to evaluate what they have purchased.

 

There are all sorts of claims about "cowboys" and "amateurs" using inferior software incompetently, and not knowing how to do proceed in a "scientific" fashion. It is quite comical, for anyone who has a few minutes to waste reading those (currently) 18 pages.

 

This pretty much shows the mentality of some vendors. Rather than address the demonstrably real and repeated problems, they would rather publicly libel and attempt to intimidate anyone who calls it into question.

[Julf]

There are all sorts of claims about "cowboys" and "amateurs" using inferior software incompetently, and not knowing how to do proceed in a "scientific" fashion. It is quite comical, for anyone who has a few minutes to waste reading those (currently) 18 pages.

 

And it is funny how uncomfortable questions are just left unanswered...

[mav52]

Chris can I recommend a free Acoustic Measurement and Analysis (REW) software to your list.

 

REW is a Java application for measuring room acoustics and analysing room and loudspeaker responses. It includes tools for generating test signals; measuring SPL and impedance; measuring frequency and impulse responses; generating phase, group delay and spectral decay plots, waterfalls, spectrograms and energy-time curves; generating real time analyzer (RTA) plots; calculating reverberation times; calculating Thiele-Small parameters; determining the frequencies and decay times of modal resonances; displaying equalizer responses and automatically adjusting the settings of parametric equalizers to counter the effects of room modes and adjust responses to match a target curve.

[Julf]

Chris can I recommend a free Acoustic Measurement and Analysis (REW) software to your list.

 

REW is a Java application for measuring room acoustics and analysing room and loudspeaker responses. It includes tools for generating test signals; measuring SPL and impedance; measuring frequency and impulse responses; generating phase, group delay and spectral decay plots, waterfalls, spectrograms and energy-time curves; generating real time analyzer (RTA) plots; calculating reverberation times; calculating Thiele-Small parameters; determining the frequencies and decay times of modal resonances; displaying equalizer responses and automatically adjusting the settings of parametric equalizers to counter the effects of room modes and adjust responses to match a target curve.

 

I think thread is abou analyzing recordings, not rooms or loudspeakers....

[mav52]

Oh thanks Julf

[REShaman]

Paul McCartney: Ram Deluxe Set Max conversion from Flac to AIFF. Thoroughly enjoying this release. I am purposely playing each disc/DL separately for the SQ etc. The two bonus tracks are very good. But then I am a fan of McCartney's music. Certain discs in the Deluxe set are only available as physical discs (the discs are in the mail). The bonus tracks are DLs only at 44.1/16. The Deluxe Set provides a download card for the 96/24. Don't know the reason for having to use the card. Purchasing the Deluxe Edition from McCartney's website added an early DL opportunity for almost the entire set in advance, but apparently, to my surprise, not at 96/24. Pourquoi? The DLs which include both audio and video are highly entertaining. The SQ is admirable. Excellent? Maybe not A+. B+ is good. Waiting for the 96/24 files for DLing before I say more. Music is excellent at any resolution.

Enjoying the past in the present,

Richard

[chriss71]

How can I compare different EQ choices by different masterings? Is this possible with Adobe Audition?

[Julf]

How can I compare different EQ choices by different masterings? Is this possible with Adobe Audition?

 

Any software that shows you a spectrum plot will allow you to do that on a rough level, but interpreting the plot takes a bit of experience.

[elcorso]

How can I compare different EQ choices by different masterings? Is this possible with Adobe Audition?

 

Please remember that the ear/brain 'system' is much faster than the sight/brain 'system'.

 

Then, at the end, it will be by listening test.

 

Roch

[Julf]

Then, at the end, it will be by listening test.

 

One method will not rule out another...

[elcorso]

One method will not rule out another...

 

I don't know.

 

Since I own a very sophisticated Real Time Analyzer, with an excellent calibrated mic, et al... that I don't use anymore: When upon the Analyzer everything is 'perfect' in my music room, I got the worst SQ. The same test made with some musician friends finished with the same conclusions...

 

Roch

[Julf]

Since I own a very sophisticated Real Time Analyzer, with an excellent calibrated mic, et al... that I don't use anymore: When upon the Analyzer everything is 'perfect' in my music room, I got the worst SQ. The same test made with some musician friends finished with the same conclusions...

 

That is interesting - what and how did you adjust the response to get the "perfect" analyzer result? And what was "perfect"?

 

In any case, I am not sure this is relevant to comparing different EQ choices by different masterings...

[elcorso]

That is interesting - what and how did you adjust the response to get the "perfect" analyzer result? And what was "perfect"?

 

In any case, I am not sure this is relevant to comparing different EQ choices by different masterings...

 

The analyzer has a LED display, but has also an PC & oscilloscope output. There is software that shows you what would be an 'perfect' (or close) output.

 

If this not relevant showing different EQ choices from different masterings I wonder which one could be... By looking at the spectrographs only?

 

Roch

[Julf]

The analyzer has a LED display, but has also an PC & oscilloscope output.

 

Well, yes, but what does the output actually show? And what did you do to change what is shown?

 

There is software that shows you what would be an 'perfect' (or close) output.

 

And what is the definition of "perfect" that the software uses? And what do you change to make the output be closer to "perfect"?

 

If this not relevant showing different EQ choices from different masterings I wonder which one could be... By looking at the spectrographs only?

 

I don't think using a real time analyzer to make things "perfect" in your listening room has very much at all to do with showing different EQ choices from different masterings, but perhaps you can explain what you think the relevance is?