"is lossless really lossless" part N

Topic: "is lossless really lossless" part N (Read 7431 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.

"is lossless really lossless" part N

2005-10-07 01:36:41

[span style='font-size:7pt;line-height:100%']From: i must be insane, listening tests[/span]

Quote

In order of importance for detecting artifacts(1 highest, 5 lowest):

1. training (probably amounts to more than 50%!)
2. hearing-abilities
3. how long you have been awake
4. noisefloor in your room (in case you use speakers)
5. equipment

As you can see, the equipment is the least important factor for detecting lossy-encoding-artifacts. I'm not making this up: some of the best encoders were tuned on laptop-soundcards with cheap headphones, if i remember right. Good studio-monitors may help a bit, but how many listeners own these?

- Lyx
[a href="index.php?act=findpost&pid=330766"][{POST_SNAPBACK}][/a]

Hi but with all due respect, what are you basing these "importances" on?

I agree that with an MP 3 player, We have IHP 140, Shuffle nad nano, I also use a pair of E2cs and a shelbrook labs head amp, but hese are for travelling, and the differences are negligible, however there is a difference, when listening on my main system, the difference between MP3 and AAC compared to CD equivelent is huge, the lossless formats seem to curtail the top end, and sound mushy in the mid bass, they also don't go anything like as low.

"is lossless really lossless" part N

Reply #1 – 2005-10-07 01:40:11

Quote

the lossless formats seem to curtail the top end, and sound mushy in the mid bass, they also don't go anything like as low.
[a href="index.php?act=findpost&pid=332248"][{POST_SNAPBACK}][/a]

Seem is the operative word. Either that or you just don't understand the definition of lossless.

"is lossless really lossless" part N

Reply #2 – 2005-10-07 01:46:17

Quote

Quote
the lossless formats seem to curtail the top end, and sound mushy in the mid bass, they also don't go anything like as low.
[a href="index.php?act=findpost&pid=332248"][{POST_SNAPBACK}][/a]

Seem is the operative word. Either that or you just don't understand the definition of lossless.
[a href="index.php?act=findpost&pid=332249"][{POST_SNAPBACK}][/a]

perhaps you just don't have equipment that can resolve enough detail to notice, I'm British (Scottish actually) using the word seem is more polite and less antgonistic, but there definately is a huge difference, overall it's duller, less dynamic range too, and it gets worse much worse below 192. and thats with AAC on a nano into main rig, what exactly are you listening through to arrive at these conclusions, may I ask?

"is lossless really lossless" part N

Reply #3 – 2005-10-07 01:56:08

Lossless is lossless, this is mathematically proven. Numbers don't lie, human ears can.

"is lossless really lossless" part N

Reply #4 – 2005-10-07 02:34:59

Quote

Quote
Quote
the lossless formats seem to curtail the top end, and sound mushy in the mid bass, they also don't go anything like as low.
[a href="index.php?act=findpost&pid=332248"][{POST_SNAPBACK}][/a]

Seem is the operative word. Either that or you just don't understand the definition of lossless.
[a href="index.php?act=findpost&pid=332249"][{POST_SNAPBACK}][/a]

perhaps you just don't have equipment that can resolve enough detail to notice, I'm British (Scottish actually) using the word seem is more polite and less antgonistic, but there definately is a huge difference, overall it's duller, less dynamic range too, and it gets worse much worse below 192. and thats with AAC on a nano into main rig, what exactly are you listening through to arrive at these conclusions, may I ask?
[a href="index.php?act=findpost&pid=332252"][{POST_SNAPBACK}][/a]

I think you are talking about LOSSY codecs. As is mentioned, LOSSLESS is LOSSLESS.

"is lossless really lossless" part N

Reply #5 – 2005-10-07 02:48:16

Quote

Quote
Quote
the lossless formats seem to curtail the top end, and sound mushy in the mid bass, they also don't go anything like as low.
[a href="index.php?act=findpost&pid=332248"][{POST_SNAPBACK}][/a]

Seem is the operative word. Either that or you just don't understand the definition of lossless.
[a href="index.php?act=findpost&pid=332249"][{POST_SNAPBACK}][/a]

perhaps you just don't have equipment that can resolve enough detail to notice, I'm British (Scottish actually) using the word seem is more polite and less antgonistic, but there definately is a huge difference, overall it's duller, less dynamic range too, and it gets worse much worse below 192. and thats with AAC on a nano into main rig, what exactly are you listening through to arrive at these conclusions, may I ask?
[a href="index.php?act=findpost&pid=332252"][{POST_SNAPBACK}][/a]

I think you should reread your posts. I bolded the important part.

"is lossless really lossless" part N

Reply #6 – 2005-10-07 03:11:56

Quote

Quote
In order of importance for detecting artifacts(1 highest, 5 lowest):
1. training (probably amounts to more than 50%!)
2. hearing-abilities
3. how long you have been awake
4. noisefloor in your room (in case you use speakers)
5. equipment
As you can see, the equipment is the least important factor for detecting lossy-encoding-artifacts. I'm not making this up: some of the best encoders were tuned on laptop-soundcards with cheap headphones, if i remember right. Good studio-monitors may help a bit, but how many listeners own these? [a href="index.php?act=findpost&pid=330766"][{POST_SNAPBACK}][/a]

Hi but with all due respect, what are you basing these "importances" on? [a href="index.php?act=findpost&pid=332248"][{POST_SNAPBACK}][/a]

Probably on what factors are statistically most likely to matter to whether or not a listener can notice artifacts, eh?

edit: Lyx, your consistent sarcasm is rubbing off on me. Shame on you!

"is lossless really lossless" part N

Reply #7 – 2005-10-07 03:20:40

Quote

I must be nuts. I can listen to a song encoded with version 3.20 of the nero AAC encodeer dll at 320 and at 128 - and while i can feel a difference, it seems like warmth... i canno tpass a listening test - I don't get it.. listening thorugh, t seem stha tthings are more chatoic and mixed.. less definition at 128 - but it's not that tangible to me.. [a href="index.php?act=findpost&pid=330537"][{POST_SNAPBACK}][/a]

Something that no one has mentioned thus far... you're describing what is commonly known as a placebo effect. That is, you know that there is a difference between two files (original wav file and a compressed/encoded file), and so you perceive a difference. However, when you don't know which file is which, you can't tell a difference. Others have pointed out that training may help you to distinguish artifacts and tell an encoded file apart from the original. But that would be above and beyond what your ears + brain can recognize now. Perhaps your hearing has degraded, but you probably just thought yourself into thinking that you heard a difference.
Anyway, congrats for being willing to take a listening test to see if you actually can hear anything when you don't already know what you think you're going to hear

Finally, those of you who think that lossless compression (despite the name) somehow changes the sound... rest assured that you'll never, ever be able to blind-test distinguish them, no matter how good your hearing is. Neither could a dog, or a whale, or a computer... Which is just to reiterate that there is absolutely no difference between an uncompressed wav file and that same file losslessly compressed.

"is lossless really lossless" part N

Reply #8 – 2005-10-07 05:03:19

Quote

...or a whale, or a computer...
[a href="index.php?act=findpost&pid=332267"][{POST_SNAPBACK}][/a]

How do you convince a whale to take a blind ABX test?
How do you even get the ABX software and the whale in the same room?
And most important: Where would you get the huge headphones?

"is lossless really lossless" part N

Reply #9 – 2005-10-07 05:44:36

Quote

Quote
Quote
Quote
the lossless formats seem to curtail the top end, and sound mushy in the mid bass, they also don't go anything like as low.
[a href="index.php?act=findpost&pid=332248"][{POST_SNAPBACK}][/a]

Seem is the operative word. Either that or you just don't understand the definition of lossless.
[a href="index.php?act=findpost&pid=332249"][{POST_SNAPBACK}][/a]

perhaps you just don't have equipment that can resolve enough detail to notice, I'm British (Scottish actually) using the word seem is more polite and less antgonistic, but there definately is a huge difference, overall it's duller, less dynamic range too, and it gets worse much worse below 192. and thats with AAC on a nano into main rig, what exactly are you listening through to arrive at these conclusions, may I ask?
[a href="index.php?act=findpost&pid=332252"][{POST_SNAPBACK}][/a]

I think you should reread your posts. I bolded the important part.
[a href="index.php?act=findpost&pid=332262"][{POST_SNAPBACK}][/a]

1) Lossless is lossless. It decodes to the exact same wav that you fed the encoder, hence the term losseless. There is no loss. ABX it, do a mathematical analysis, do what ever you want, there is no loss.

2) AAC is not lossless, it is a lossy encoder that tries to sound close to the input source but something must be lost. Most lossy encoders are tuned to sound transparent (non-abxable) to the source. If you can consistantly abx lossy encoders, perhaps you should consider participating in listening tests.

"is lossless really lossless" part N

Reply #10 – 2005-10-07 08:55:04

Quote

Quote
Quote
Quote
Quote
the lossless formats seem to curtail the top end, and sound mushy in the mid bass, they also don't go anything like as low.
[a href="index.php?act=findpost&pid=332248"][{POST_SNAPBACK}][/a]

Seem is the operative word. Either that or you just don't understand the definition of lossless.
[a href="index.php?act=findpost&pid=332249"][{POST_SNAPBACK}][/a]

perhaps you just don't have equipment that can resolve enough detail to notice, I'm British (Scottish actually) using the word seem is more polite and less antgonistic, but there definately is a huge difference, overall it's duller, less dynamic range too, and it gets worse much worse below 192. and thats with AAC on a nano into main rig, what exactly are you listening through to arrive at these conclusions, may I ask?
[a href="index.php?act=findpost&pid=332252"][{POST_SNAPBACK}][/a]

I think you should reread your posts. I bolded the important part.
[a href="index.php?act=findpost&pid=332262"][{POST_SNAPBACK}][/a]

1) Lossless is lossless. It decodes to the exact same wav that you fed the encoder, hence the term losseless. There is no loss. ABX it, do a mathematical analysis, do what ever you want, there is no loss.

2) AAC is not lossless, it is a lossy encoder that tries to sound close to the input source but something must be lost. Most lossy encoders are tuned to sound transparent (non-abxable) to the source. If you can consistantly abx lossy encoders, perhaps you should consider participating in listening tests.
[a href="index.php?act=findpost&pid=332281"][{POST_SNAPBACK}][/a]

Not in the habit of doing test, I prefer listening to music, as for lossy codecs, easy, even in isolation, but then I have a very revealing system, I will connect up the computer one day, and try an abx of a lossless file, but I'm sure the difference will be there, perhaps it's the resolution of the playback system, I already asked the question, what kit do you do your abx tests on? I'm sure if I abx a lossy file on my computer say AAC at 192, I could probably not tell difference blind, on my main rig however the difference is obvious, you could tell from another room.

So to all you abx lossless guys, what you comparing the differnces on?

"is lossless really lossless" part N

Reply #11 – 2005-10-07 09:05:31

If you are still claiming that you can hear a difference in losslessly compresed files (not lossy, such as AAC), then you still don't know what you are talking about.

Lossless means that has no loss. Hence the resolution of the listening sytem has no relevance at all. Period.

"is lossless really lossless" part N

Reply #12 – 2005-10-07 10:17:47

Quote

If you are still claiming that you can hear a difference in losslessly compresed files (not lossy, such as AAC), then you still don't know what you are talking about.

Lossless means that has no loss. Hence the resolution of the listening sytem has no relevance at all. Period.
[a href="index.php?act=findpost&pid=332311"][{POST_SNAPBACK}][/a]

Why not humour me, go on what you listening to it on?

saying the same thing over and over again does not make it true, back up your claims, or answer the question, wav files have say 75meg, same file lossless is less, so somethings different, as i have already eluded, in a lower res system the differences you can hear (if any) will be negligible, on a hi res system it's obvious.

you guys need to get out more I feel.

If you give me a list of the equipment you compare it on I'll answer back, but if you continue this nonsense that you are right as I don't understand, (I do, it's you guys that are blinkered) then my initial impressions about you guys will be proved (to me) right.

"is lossless really lossless" part N

Reply #13 – 2005-10-07 10:29:57

or answer the question, wav files have say 75meg, same file lossless is less, so somethings different, as i have already eluded, in a lower res system the differences you can hear (if any) will be negligible, on a hi res system it's obvious.

Guy, make up your reading, homework, or you are shortly to be banned from this scientific place.
some explanation extra for you:
Of course a wav has eg. 75 MB, the Lossless Flac or Wavpack file of this compressed wave has only ca. 40 MB sizes, as example, eg. true for lots of classical music, which compresses losslessly very well.

So, you are arguing, that then half of the content of the original file would be missing ?!
Lol.
The decoder of your player decompresses the packed Lossless Wavpack or flac file back to the original big wave file.

You can convince yourself, by trying it out on your own PC.
Make a bit to bit compare of the original wav file with the "wav->Loslsess-compressed-file->decoded-wav-file"
Result: identical,
or your PC is flawed, like unstable memory, psu, HD etc.

The soundcard etc. has no influence to the Lossless files (regarding this discussion), as the wav files can be burned to CD again, and played on highest equipment for 100k dollars, and the copy of the original CD will be bit perfect.

"is lossless really lossless" part N

Reply #14 – 2005-10-07 10:52:19

Quote

Quote
If you are still claiming that you can hear a difference in losslessly compresed files (not lossy, such as AAC), then you still don't know what you are talking about.

Lossless means that has no loss. Hence the resolution of the listening sytem has no relevance at all. Period.
[a href="index.php?act=findpost&pid=332311"][{POST_SNAPBACK}][/a]

Why not humour me, go on what you listening to it on?

saying the same thing over and over again does not make it true, back up your claims, or answer the question, wav files have say 75meg, same file lossless is less, so somethings different, as i have already eluded, in a lower res system the differences you can hear (if any) will be negligible, on a hi res system it's obvious.

you guys need to get out more I feel.

If you give me a list of the equipment you compare it on I'll answer back, but if you continue this nonsense that you are right as I don't understand, (I do, it's you guys that are blinkered) then my initial impressions about you guys will be proved (to me) right.
[a href="index.php?act=findpost&pid=332321"][{POST_SNAPBACK}][/a]

So, according to your theory - if you have, say, a Word file - of 10 megs, and you zip it - end up with the file of 2 megs - something has been lost? right?

WRONG - losless codecs compress music data in pretty much similar way like file compressors compress data - during decompression, the PCM signal is exactly the same as in the original file - without a single bit difference.

Is it more clear now?

"is lossless really lossless" part N

Reply #15 – 2005-10-07 11:01:08

Some more reading for Akid: Is a process really lossless.

To answer your question Akid; Audiophile 24/96 > Arcam A85 > Monitor Audio Silver S8s. I can assure you, this information is completely irrelevant with regards to your claims.

"is lossless really lossless" part N

Reply #16 – 2005-10-07 11:03:32

Hi Akid

Quote

saying the same thing over and over again does not make it true, back up your claims, or answer the question,

No. It is up to you to backup your claims, as you claim what has been proven wrong several times.

You claim that there is a loss in lossless compression. There has been mathematical proof, that there is no loss. So it is up to you to proof this wrong.

Quote

wav files have say 75meg, same file lossless is less, so somethings different,

Take a word-doc, take a look at its filesize. Zip it - what means kompressing it losslessly - and have again a look at its filesize. Unzip it again and open it. Have text and figures changed? Has texte disapeared? Is something missing?

Do this with an excel-sheet. Does the balance alter, do black figures turn into red ones? No, so why do you claim a loss in lossless compression?

Quote

as i have already eluded, in a lower res system the differences you can hear (if any) will be negligible, on a hi res system it's obvious.

This is just plain nonsense. Believe me, you are not the only one with high-end gear.

Quote

If you give me a list of the equipment you compare it on

Professional monitors, DSP-Speakers, professional soundcard etc.

Quote

I'll answer back, but if you continue this nonsense that you are right as I don't understand, (I do, it's you guys that are blinkered) then my initial impressions about you guys will be proved (to me) right.

Mods and admins, please bann this ignorant fool.

cu
Etienne

"is lossless really lossless" part N

Reply #17 – 2005-10-07 11:04:36

On a side note:

I see this discussion reoccurring every few weeks ("i can ABX lossless from the original"). Isn't there (or shouldn't there be as I couldn't find it this soon in the lossless codecs forum) a sticky that explains why lossless and lossless, so these claims can be countered right away instead of making a subtopic out if it in every topic it emerges.

Arguments like these:

Quote

wav files have say 75meg, same file lossless is less,so somethings

can easily be countered by giving examples of of ZIP and RAR compression, which are also lossless (I have never seen my zipped word documents been altered after extracting them).

Methods for doing bit-by-bit comparisons of original and the uncompressed lossless file can also be described to convince more stubborn people. I'll gladly assist in making such a sticky, cause this is starting to annoy me a bit.

I know this will not stop people from claiming that they can hear differences, but at least we can redirect them to something to read so it doesn't derail the original thread anymore. I don't see the point in having the same arguments over and over again.

edit: i see Ettiene and me are on the same line with the word document

edit2: now why is evereux's thread not stickied?

"is lossless really lossless" part N

Reply #18 – 2005-10-07 11:22:12

Akid... It's hard to discuss with somebody that is unable to grasp even the basics of the issue at discussion....

Anyway, I'll try. Do you know that is a zip file? Is a losslessly compressed file. It can contain audio, video, text, software, anything that can go in a computer file. It reduces the size of those, however no information from them is lost when done. It's lossless. It all can be recovered. Bit-perfect.

Haven't you ever used a zip file? Does the fact that it takes less space make you think something is lost in the process? The same goes with audio.

Edit: oh, you people beat it to me.

"is lossless really lossless" part N

Reply #19 – 2005-10-07 13:34:57

this Akid guy (or, small possibility, girl) is a pretty effective troll. Got about 20 responses all saying the same thing...

Notice