Buy/Sell/Trade :  Talk
This is the place where Stones fans can advertise anything for sale, wanted, trade or whatever, from fan to fan. Advertisements are for free.
To see the old ads go here

Previous page Next page First page IORR home

For information about how to use this forum please check out forum help and policies.

For spectrum analyses, which is the best? EAC or Audacity?
Posted by: coowouters ()
Date: August 4, 2015 19:45

I did a spectrum analysis of the same file with Audacity and EAC.
Which one is the best?

EAC:



Audacity:



.

Chris from Belgium


Re: For spectrum analyses, which is the best? EAC or Audacity?
Posted by: selgas ()
Date: August 4, 2015 20:33

Hi Chris, as far as I know, the Spectrum Analysis are a series of math functions that you finally see with a graphic representation (so you can "see" the results); so that means the result of the math fuctions should be the same whatever program you use.

Anyway, I do prefer Audacity's graphic representation as it is quite clear where does the spectrum ends in Hz (that way you can check if a source is loussy or not, but not all of the times!)

Hope I'm answering your question!

Re: For spectrum analyses, which is the best? EAC or Audacity?
Posted by: coowouters ()
Date: August 4, 2015 20:54

Quote
selgas

Hope I'm answering your question!

I also prefer the graphic representation of Audacity more than the one from EAC.
And I suppose it's reliable, no?

As for the graphics, if you get something like the following, you're pretty certain it's lossy:

Same file (converted from mp3 to wav, don't pay attention to the file name in the last picture) with Audacity and EAC:

Audacity:


EAC:


With EAC, you better use the "spectral view" to make sure:


Chris from Belgium


Re: For spectrum analyses, which is the best? EAC or Audacity?
Posted by: selgas ()
Date: August 4, 2015 21:05

Altough it does not work 100%, a fast check for loussy sources is the final Hz in Audacity:
-> around/over 20000 it's usually lossless
-> around 14000 to 18000 it's loussy.

Re: For spectrum analyses, which is the best? EAC or Audacity?
Posted by: stonesmuziekfan ()
Date: August 6, 2015 13:05

I was thinking of posting the same thread, but first was consulting "the experts"

Quote
Q&A ErikSnow
Hi Erik,
I have been posting my entire Mick Taylor collection at IORR.
Before I did all files were checked by Peter (Poetry) using Audacity.
The outcome with Audacity for files 2000 until today I thing are doubtfull for those files mainly come from torrents, so I did un some EAC analyses myself.
I see some great differences.
I don't know if you are an expert on this matter, but if so, what is to your opinion the most trustworthy software to check for lossy file, for I hate to poison the pool.
Thanks
Jaap

Erik replied
Hi Jaap
The best source is the EAC analyses *if* you are good at using that software/watching the analysis
But I'm not good at explaining in English, and besides I haven't been torrenting much or using EAC in about 4 years now, so: please ask Jimmyshelter; he is knowledgeable
Best
Erik

Than I asked Jimmyshelter the same question

Quote
Answer by Jimmyshelter
I don’t think there is a perfect answer, or the right software really.
The graph drawn by EAC (Spectral) is useful, I get a similar graph in WaveLab.
The things to look for there are a fairly smooth colour going up to the top, black at the top indicates that there is no audio there.
‘Blocky’ or ‘Lego Block’ patterns are not good, also black ‘holes’, they are all a sign of MPEG compression.
(I’m not sure the ‘FFT 2048 Hanning’ settings matter too much.)
Cut-off above about 16 kHz is also a sign of MPEG compression.

See my Forum post on EL: [electricladyland.synology.me] So if you have two files to compare;
File A shows a smooth colour change right to the top and no cut-off.
File B is cut-off steeply at about 16 kHz and shows a black band at the top of the Spectrum graph and some black holes.
So File B is clearly a lossy (MPEG probably) copy of File A (if both files should be the same).
If the 2 files are not expected to be the same though then we can say:
File A is probably lossless because it has a full spectrum and no signs of MPEG.
It is still not so easy to say that File B is lossy though unless we have something to compare it to.
If it has a steep cut-off at about 16 kHz and has a blocky appearance then it is PROBABLY lossy, but there may be other reasons.

Remastering (even just reducing the hiss) can result in the top-end being cut-off and this is where it becomes difficult.
There are also different types of lossy compression, not all as easy to detect as MPEG.
Some look no different to lossless.

If you look at the files in a WAV editor, like Audacity, look at the ends of the waveforms
although it is possible to cut the ends and convert to flac then that bit of evidence will be lost.

We have just had a debate with Poetry, among others, about the Stones Coventry ’71.

He regards anything with no top-end as ‘Lossy’, but ‘Lossy’ really has a specific meaning concerning compression techniques.
It also depends on what the actual source was, if the original source was lossy (ie, MP3 recorder, MD, etc) then that would be the ‘best available source’.

From 2000 on you would have thought that they would be digital recordings, maybe they were recorded in a lossy format?
Without having something to compare then it is often very difficult, if even possible, to be certain.

What we would say on EL would be that unless there is a know better source, then put your concerns in the description (ie, “spectral analysis shows possibly lossy source but no better source known” or something)
Then it would run as the best available source on the understanding that if a know source turns up it would be replaced.

I assume from your questions that you are not aware of any better versions existing. It may be that there aren’t any.
There is a lot said about Lossy/Lossless, and a lot more made of it than is healthy I think sometimes.
Some people seem to be obsessed by it, and it seems that the first thing they do is to look rather than listen.

I’m sure you were hoping for a more precise answer, I’m afraid I don’t have one.
It is not always as easy as people seem to think.

So here we are .....

Hope this all helps

Re: For spectrum analyses, which is the best? EAC or Audacity?
Posted by: selgas ()
Date: August 6, 2015 14:03

I thik Chris was having an extra problem as the sources he vas checking were "from the radio", and that kind of recordings do have lot of frequencies compression...

Re: For spectrum analyses, which is the best? EAC or Audacity?
Posted by: coowouters ()
Date: August 6, 2015 14:26

Quote
Answer by Jimmyshelter

There is a lot said about Lossy/Lossless, and a lot more made of it than is healthy I think sometimes.
Some people seem to be obsessed by it, and it seems that the first thing they do is to look rather than listen.

Amen to that!!!!
My hearing isn't what it was anymore (thank you Rolling Stones and shooting assault rifles without proper hearing protection winking smiley )

.

Chris from Belgium


Re: For spectrum analyses, which is the best? EAC or Audacity?
Posted by: alimente ()
Date: August 6, 2015 18:48

Quote
coowouters
Quote
Answer by Jimmyshelter

There is a lot said about Lossy/Lossless, and a lot more made of it than is healthy I think sometimes.
Some people seem to be obsessed by it, and it seems that the first thing they do is to look rather than listen.

Amen to that!!!!
My hearing isn't what it was anymore (thank you Rolling Stones and shooting assault rifles without proper hearing protection winking smiley )

.

Well, don't shoot me - I'm only the piano player, Chris!

Cheers!



Sorry, only registered users may post in this forum.

Online Users

Guests: 1831
Record Number of Users: 206 on June 1, 2022 23:50
Record Number of Guests: 9627 on January 2, 2024 23:10

Previous page Next page First page IORR home