How to tell if your FLAC files are 'really' FLAC

[creatip123]

FLAC files are all the craze on audio lovers nowadays. FLAC (Free Lossless Audio Compression) is supposed to give a size compression to original PCM uncompressed (wave), without any loss in quality and frequency range of the original file. 

 

Now how do you know the FLAC files you have, that you don't encode yourself (maybe downloaded/bought online) are 'real' FLAC, meaning they were encoded straight from a good full frequency range source? 

 

I found an article on the net, on how to test the 'quality' of the FLAC files you have. I'm sure many of you know about this already, but what the heck.....

 

- First you need to download a freeware program called 'Spectro', and install it. It's light and totally free.

- Run the program, open a file you want to test. It will show vital informations of the files

 

For example, I got these audio files from a website that provide various file formats for a particular song for download. So I'm assuming they do the encoding of the different formats from the same master source.

 

 

This is the result of the FLAC file testing. You can see the spectrometer goes all the way up to full 22KHz. Because the sample rate of this file is 44khz only, so full spectrometer at 22khz (44 divided by 2) is very good. You can also see other informations on the file, like the compression ratio, original size, etc.

 

 

This is a result of the 320kbps MP3 version of the same song. You can see the spectrometer got decapitated a little short of 22khz. The reason is because MP3 encoding uses a lossy compression, so you can see the loss there in the spectrometer and also in the frequency graph in the lower left.

 

 

You might have guessed, the 128kbps version got decapitated much more brutally there. 

 

You can encode a lossy, low quality MP3 (say 128kbps like above) to FLAC, but of course you won't get the full spectrum that way. 

 

 

This is the reading of a FLAC file I converted from the 128kbps MP3 source. It's the exact same with the picture before it. So although it's a FLAC file, but because it's encoded from a lossy format, it doesn't have the benefit of the lossless compression. Funny thing is, the source file (128kbps MP3) is 3.8MB, while the FLAC file is 33MB. So I was actually wasting almost 10 times of storage space, for the same exact quality. In other words, totally overkill.

 

So what's the meaning of all this? This is a way to tell if your FLAC files got encoded from a full frequency range source, or from a lossy source, like the picture above. The spectro reveals the 'true face' of the FLAC files you have. 

 

 

Before I get death threats, I want to clarify: this is just a technical testing method for the digital audio files you have. It's more intended for 'purists' who are concerned about technical features. In reality I doubt there are many people in the world who can actually hear the difference of a full spectrum FLAC and a 320kbps MP3. As you can see from the pictures above, the frequency cut-off of the 320kbps MP3 is very little compared to the full spectrum FLAC version. Heck, sometimes I can't even spot  128kbps MP3s. My advice is, try this (if you want to) just for fun only, and then get back to enjoying your music.

[Fred Castellum]

Very interesting thanks for letting us know. 

[SirRoderick]

Hah....if you need to do that kind of analysis to figure out of it's a true FLAC what's the point of the FLACs in the first place? I find it extremely hard to hear a difference between them and 320kbps MP3s and I do use fairly good sound equipment.

[Zeke]

Hah....if you need to do that kind of analysis to figure out of it's a true FLAC what's the point of the FLACs in the first place? I find it extremely hard to hear a difference between them and 320kbps MP3s and I do use fairly good sound equipment.

But the chart says...

[Enderman]

So by 'lossy' and 'low quality' you mean it doesn't reach 22khz?

If by listening to 128kbps music i dont get to hear a 22khz sound i dont really care.

Its not like you need to hear those barely audible squeaky noises...

From what I can tell in the images, the most important part of the music is still the same throughout all the compressions. (0~18khz)

[SirRoderick]

But the chart says...

Hey I'm sure some people can hear the difference more than I can. For me though it's taking up a LOT of hard drive space for no real reason.

[Zeke]

Hey I'm sure some people can hear the difference more than I can. For me though it's taking up a LOT of hard drive space for no real reason.

Yeah, I actually think the same thing. For me there's no point for going beyond 320 kbps MP3 (or below 128, lower bitrate is noticeably worse). 

 

Still, the post is entertaining, I've never considered "x-raying" audio files.

[SirRoderick]

Yeah, I actually think the same thing. For me there's no point for going beyond 320 kbps MP3 (or below 128, lower bitrate is noticeably worse). 

 

Still, the post is entertaining, I've never considered "x-raying" audio files.

Oh yeah it's a lovely procedure ^^

[brownninja97]

250+kbps or just flac, ive got the space for it and if it becomes a problem then ill start compressing.

[Mooshi]

I find my ears to be more reliable than a bunch of fancy graphs.

 

When the audio is clear and doesn't distort at high volume, I am pleased. If things sound really compressed like the audio is being dragged through mud, I am not. The format doesn't matter, clarity does.

[Tech_Dreamer]

This is Actually a nice Post! Thanks 

 

Edit:

 

Oh, reason i liked the post is that , i can tell the diference between mp3 128kbps vs 320kbps vs 900-1200kbps flac somehow. But people i'm with are used to normalised/boosted mp3 audios & saying that it's better , Flac/wav is just a waste of space,, i'm somewhat fine tuned to details in music or movie audio.especially the really deep rumbling lows i can somehow feel it. i'm not a fan of wav files , which are the truest form of raw data file. but still i love high bitrate audio.

 

 

 

  a ninja can see sounds

[BoredErica]

Hey I'm sure some people can hear the difference more than I can. For me though it's taking up a LOT of hard drive space for no real reason.

A lot?

Hardly, unless we're talking Ipods or something.

 

So by 'lossy' and 'low quality' you mean it doesn't reach 22khz?

If by listening to 128kbps music i dont get to hear a 22khz sound i dont really care.

Its not like you need to hear those barely audible squeaky noises...

From what I can tell in the images, the most important part of the music is still the same throughout all the compressions. (0~18khz)

No, that's not what it means.

Frequencies we can hear are also distorted. And older recordings that are FLAC may not have data in those frequency ranges, giving a false negative for the test.

 

It's not a race to see which file has higher frequency data.

 

 

 

Here is a thread about this:

http://www.head-fi.org/t/442888/how-to-tell-between-a-real-flac-and-a-fake-flac

[Recon-UK]

No.

[SirRoderick]

 

A lot?

Hardly, unless we're talking Ipods or something.

 

Ehhh, yeah, a massive amount. My folder of the ENTIRE collection of Beethoven's Piano sonatas in MP3 320kbps (106 files) is less than half as large as my Tool collection in FLAC (84 files) That is an enormous difference in storage capacity. If I turned all of my FLAC files into MP3s I would save over a hundred GBs of space on this TB drive, so about 10% of the total capacity of my hard drive is spent on storing FLACS over MP3's.That is too much for something that doesn't actually improve my listening experience.

[TheDigitalRealm]

I just buy CDs and do the encoding myself.. I prefer a hard copy as well

[Lays]

 

q

 

 

 

 

IMO it has less to do with size/file type and more to do with the mastering in the first place, some songs no matter what format have really bad backround noise, from poor mastering.

 

I can hear the difference in files most of the time but the difference is quite minuscule honestly.

 

As long as it sounds good to me I don't bother worrying to much about the file type etc. I listen to spotify premium half the time anyways so it's whatever lol

 

Maybe if I had $40,000 worth of snakeoil products It'd be more clear, but I'm fine with my he-500's and o2 odac lol

[creatip123]

Erghhhh....my quote box in the original post seemed to have escaped people's attention.....

[Enderman]

 

A lot?

Hardly, unless we're talking Ipods or something.

 

No, that's not what it means.

Frequencies we can hear are also distorted. And older recordings that are FLAC may not have data in those frequency ranges, giving a false negative for the test.

 

It's not a race to see which file has higher frequency data.

 

 

 

Here is a thread about this:

http://www.head-fi.org/t/442888/how-to-tell-between-a-real-flac-and-a-fake-flac

 

Do you have any pictures/visual demonstrations of how the audi is distorted in 128kbps vs 320kbps?

Because fro the images you posted they seem pretty much identical...

I'm not denying the fact that 128kbps is much worse quality, but i would like to see how

[BoredErica]

IMO it has less to do with size/file type and more to do with the mastering in the first place, some songs no matter what format have really bad backround noise, from poor mastering.

 

I can hear the difference in files most of the time but the difference is quite minuscule honestly.

 

As long as it sounds good to me I don't bother worrying to much about the file type etc. I listen to spotify premium half the time anyways so it's whatever lol

 

Maybe if I had $40,000 worth of snakeoil products It'd be more clear, but I'm fine with my he-500's and o2 odac lol

No, the fact of the matter is that a properly done mp3 is very similar to a FLAC file to our ears. Very few people can reliably tell a difference.

Mastering matters but we don't do the mastering so sometimes it's out of our control entirely. What we can control is the file type we choose. 

 

Do you have any pictures/visual demonstrations of how the audi is distorted in 128kbps vs 320kbps?

Because fro the images you posted they seem pretty much identical...

I'm not denying the fact that 128kbps is much worse quality, but i would like to see how

No, I do not. But consider this: Nobody ever claimed, even from the people who think FLAC is a night and day difference, that the difference is due to super high or low frequencies. That was never the claim.

I don't even want to hear frequencies beyond 20khz, it wouldn't be all that pleasant to hear I think. Not a whole lot of detail up there.

 

You can do the Golden Ear Challenge (google it to find it) and there is a section dedicated to bitrates. The lower bitrate samples are not exactly missing the high end notes, they're just not as good. My vocab is lacking here but click one of the lower bitrates to hear an obvious difference. Actually, the bit rate section was the hardest section for me to do in the entire challenge. Because of that section I could not pass the Silver Ears mark. 

 

 

 

[creatip123]

You can do the Golden Ear Challenge (google it to find it) and there is a section dedicated to bitrates. The lower bitrate samples are not exactly missing the high end notes, they're just not as good. My vocab is lacking here but click one of the lower bitrates to hear an obvious difference. Actually, the bit rate section was the hardest section for me to do in the entire challenge. Because of that section I could not pass the Silver Ears mark. 

 

 

Yeah, the bitrate samples test was one of the hardest for me too. Up to the point I started searching more info on that particular test on the net. Found a forum post somewhere that the source audio they use is 16/44.1, so you gotta set the window's sample rate to that config. If it got upsampled, it's much harder to hear the differences. Also I found an unorthodox way to cheat on that bitrate samples test: I set my audio panel's reverb to the most obscure effect I could find. The distorted artifacts are easier to hear that way.

 

After that, you got 1 more ultimate test, in golden ears, which is an 8 bands EQ setting (find which frequency band is missing or boosted). Not easy, but I graduated from the test....

[BoredErica]

Yeah, the bitrate samples test was one of the hardest for me too. Up to the point I started searching more info on that particular test on the net. Found a forum post somewhere that the source audio they use is 16/44.1, so you gotta set the window's sample rate to that config. If it got upsampled, it's much harder to hear the differences. Also I found an unorthodox way to cheat on that bitrate samples test: I set my audio panel's reverb to the most obscure effect I could find. The distorted artifacts are easier to hear that way.

 

After that, you got 1 more ultimate test, in golden ears, which is an 8 bands EQ setting (find which frequency band is missing or boosted). Not easy, but I graduated from the test....

You talking about control panel -> audio hardware -> Odac -> 16bit 44khz in shared mode?  :blink:

[creatip123]

You talking about control panel -> audio hardware -> Odac -> 16bit 44khz in shared mode?  :blink:

 

Dunno about the shared mode, but yeah, that's the setting. On general use, I just go with the highest setting available, but for that particular test, I needed to set it to the native 16/44 setting.

[BoredErica]

Mine is already at 16/44. Any higher is pointless.

 

Ehhh, yeah, a massive amount. My folder of the ENTIRE collection of Beethoven's Piano sonatas in MP3 320kbps (106 files) is less than half as large as my Tool collection in FLAC (84 files) That is an enormous difference in storage capacity. If I turned all of my FLAC files into MP3s I would save over a hundred GBs of space on this TB drive, so about 10% of the total capacity of my hard drive is spent on storing FLACS over MP3's.That is too much for something that doesn't actually improve my listening experience.

I have 279 files in my music folder. Average bit rate is 660 kbps.

 

166 FLAC files: 4.2gb

114 MP3 files: 573mb

 

I don't know how many tracks you actually have in your collection.

 

166 Flac files = 4200mb

1 Flac file = 25.3mb

 

114 Mp3 files = 573mb

1 Mp3 file = 5.02mb

 

Here I'm estimating that an average FLAC file is 5 times the size of a similar MP3 file.

F = Flac

M = MP3

 

F - M = 100,000 mb (A very generous assumption as you said over a hundred)

F = 5M

 

5M - M = 100,000 mb

M = 25,000mb

 

In order to save 100gb of space by choosing MP3 over FLAC you need to have 25gb worth of MP3 files. 

You'll need to have 5,000 5mb MP3 tracks to have 25gb of MP3 files. Therefore in order for to save 100gb of space (exactly) going MP3 vs FLAC, you need to have ~5000 MP3 tracks.

 

 

Now while it's possible that you have 5000 tracks in your library, A. I think it's more likely that you made a miscalculation but even if not, B. I don't think your case is typical, like at all. If you have that many tracks you should probably have noted your large library size at the start.

 

Or C. I made a miscalculation that invalidates my point. But the numbers look fine to me.

 

If you legitly bought 5000 tracks, another terrabyte hard drive is not a deterrent to you. At a buck a track, that's $5000. A terrabyte hard drive is $60. And even then you will only be using 125gbs of the 1 TB drive, netting you another 875gbs of free space for 1/83rd the cost of your music collection.

[xyza]

heh

[SirRoderick]

 

Mine is already at 16/44. Any higher is pointless.

 

I have 279 files in my music folder. Average bit rate is 660 kbps.

 

166 FLAC files: 4.2gb

114 MP3 files: 573mb

 

I don't know how many tracks you actually have in your collection.

 

166 Flac files = 4200mb

1 Flac file = 25.3mb

 

114 Mp3 files = 573mb

1 Mp3 file = 5.02mb

 

Here I'm estimating that an average FLAC file is 5 times the size of a similar MP3 file.

F = Flac

M = MP3

 

F - M = 100,000 mb (A very generous assumption as you said over a hundred)

F = 5M

 

5M - M = 100,000 mb

M = 25,000mb

 

In order to save 100gb of space by choosing MP3 over FLAC you need to have 25gb worth of MP3 files. 

You'll need to have 5,000 5mb MP3 tracks to have 25gb of MP3 files. Therefore in order for to save 100gb of space (exactly) going MP3 vs FLAC, you need to have ~5000 MP3 tracks.

 

 

Now while it's possible that you have 5000 tracks in your library, A. I think it's more likely that you made a miscalculation but even if not, B. I don't think your case is typical, like at all. If you have that many tracks you should probably have noted your large library size at the start.

 

Or C. I made a miscalculation that invalidates my point. But the numbers look fine to me.

 

If you legitly bought 5000 tracks, another terrabyte hard drive is not a deterrent to you. At a buck a track, that's $5000. A terrabyte hard drive is $60. And even then you will only be using 125gbs of the 1 TB drive, netting you another 875gbs of free space for 1/83rd the cost of your music collection.

 

All of which is entirely beside the point that FLACs take up a lot more space than MP3s without providing me any benefit, therefore are objectively worse for me. 5 times the space is ludicrous, some people actually USE their storage space. Arguing "oh well you might as well buy more storage equipment then!" is absolutely backwards, no I should not have to do that, I should be switching over to MP3s for free

 

And yes I do have that many tracks, it is a massive space consumption fo have so many files in FLAC. One dollar per track would be a very expensive estimate, most music is either older stuff I grabbed off of torrent sites or music from bandcamp which is often much much cheaper than that.