HolidayBuyer's Guide

TVs & Home Theaters forum

General discussion

Can someone explain Lossless

by Rollbar / January 24, 2008 3:20 AM PST

It comes up all the time and I know it's important to Jostenmeat.

What the heck is it?

Gary

Discussion is locked
You are posting a reply to: Can someone explain Lossless
The posting of advertisements, profanity, or personal attacks is prohibited. Please refer to our CNET Forums policies for details. All submitted content is subject to our Terms of Use.
Track this discussion and email me when there are updates

If you're asking for technical help, please be sure to include all your system info, including operating system, model number, and any other specifics related to the problem. Also please exercise your best judgment when posting in the forums--revealing personal information such as your e-mail address, telephone number, and address is not recommended.

You are reporting the following post: Can someone explain Lossless
This post has been flagged and will be reviewed by our staff. Thank you for helping us maintain CNET's great community.
Sorry, there was a problem flagging this post. Please try again now or at a later time.
If you believe this post is offensive or violates the CNET Forums' Usage policies, you can report it below (this will not automatically remove the post). Once reported, our moderators will be notified and the post will be reviewed.
Collapse -
well, my understanding is that
by jostenmeat / January 24, 2008 3:42 AM PST

you "lose" nothing from what is mastered in the studio. That's right, you are supposedly getting 100% of the effect as the guy who is creating the soundtrack in the studio.

For lossy algorithms, they take a LOT of information out to squeeze the audio on to a puny format (mp3, er, even a DVD).

See, they figured they could do this because a lot of the removed audio-info is not perceivable by humans anyways. Saves a lot of space without compromise, in theory. The problem is they still are taking away a lot of which we can hear.

Lossless does not mean uncompressed. You can have a compressed yet lossless audio track.

Just think of it as the best audio you can possibly get.

Collapse -
Well...
by froasier / January 27, 2008 12:34 AM PST

you probably won't get "100% of the effect as the guy who is creating the soundtrack in the studio" because his studio is probably better than your living room, and he may be working at better than CD quality.

What you will get is 100% of the effect as if you had played the uncompressed version, i.e. the original audio CD.

Most people, however, cannot tell the difference between a well-encoded mp3 and the original (especially on lower end equipment), so lossless may or may not be for you.

Collapse -
Lossless compression is a myth
by mjd420nova / January 25, 2008 12:08 PM PST

I've tested many compression schemes in a laboratory and found none to be lossless. The idea that what is removed by the mathematical formulas is not percieved by the human ear and thereby not needed is wrong. Much of what the human ear percieves is not determined by the actual content but the contents effect on the ear drum. This includes the subharmonics created by many notes beating against each other, creating content that is not present in the original recordings nor recreated by their translation from sound waves to electrical signals by microphones or other pickup devices. Again, this content is retranslated back to sound waves by amplifiers and speakers. Subtracting any amount of this content however small does affect the eventual output of the content and its perception by the human ear.

Collapse -
Compression vs. digitisation
by genotypewriter / January 25, 2008 10:31 PM PST

I don't know why you needed a lab to answer this question as you only need a PC. Compress something like a text file in ZIP or RAR formats and uncompress it. Use whatever tool you like to do this but just compare the two files to see if they're any different... and they shouldn't be because they're lossless.

With images, TIFF, PNG, BMP are lossless formats. With sound there's PCM .WAV.

It seems to me that you're talking about the loss of information in the digitisation process. This is not related to lossy or lossless concepts as the latter relate to compression. Digitisation does quantise information and this theoretically causes a loss of detail (how much depends on the sampling amounts)... that's why for example some people prefer vinyl over CDs and film photos over digital photos.

Collapse -
hold on
by froasier / January 27, 2008 12:47 AM PST

TIFF is a container that can hold uncompressed, losslessly compressed, or lossily compressed data. BMP is not compressed. WAV files are also uncompressed. PNG however, is a good example of lossless compression, as would be FLAC for audio.

Collapse -
Was referring to lossless TIFF
by genotypewriter / January 31, 2008 5:54 PM PST
In reply to: hold on

Hold on to what?

Collapse -
Lossless compression
by mjd420nova / January 28, 2008 2:59 AM PST

I work in a standards lab and had access to very accurate equipment and did the testing just to satisify my own questions. As an example, I started with just a single tone, a 1000 cycle signal, compressed it, decompressed it using a number of different formats and then measured the result. Even with just a single tone, i was amazed to find that the output was not 1000 cycles anymore, but anywhere from 990 to 998 cycles. Now grated most users would never notice the difference, but we're talking about a very complex signal input when it comes to music. Now this is a completely different problem when dealing with photos as no one will notice anything whena few pixels are missing or might be the wrong color. Data is very, very different, as the loss of even a single bit can mean diaster for an accounting program or a trajectory program for a satelite. Music is quite complex and compression will not be noticed by 90 percent of the users.

Collapse -
it sounds like
by froasier / January 27, 2008 12:53 AM PST

you are testing lossy compression methods; therefore it is unsurprising that they are... lossy.

If you tested actual "lossless" formats like FLAC (were you aware these existed?) you would find that nothing is removed from the original digital recording.

Collapse -
the nature of losslessness
by jbelck / January 25, 2008 12:59 PM PST

To impress you, at absolute zero, wire of any kind offers no resistance to the flow of electricity.
To inform you, a recording of any kind would be lossless only if what you heard was absolutely the same to your ears as a live performance.
In othe words, lossless is hype.

Collapse -
Compression artifacts
by genotypewriter / January 25, 2008 10:41 PM PST

With lossy compression, you get compression artifacts. Maybe, the devices you use now can't resolve enough detail for these to be a problem but think about in a couple of years time you might wish you had the uncompressed/lossless compressed originals. This is why MiniDV is still preferred over HDD camcorders and

Lossless in general is definitely not hype. Lossy saves a lot more space and is easier to transfer. But why do you think all studio work is done in uncompressed/lossless formats if they're really just hype? Plus if you're zipping or rar'ing your files, you'll corrupt the files if they're not lossless.

Collapse -
As an aside....
by forkboy1965 / January 26, 2008 8:37 AM PST
In reply to: Compression artifacts

As an aside...thanks for bringing forward a very good reason to stick with MiniDV-based camcorders, which I have been putting forth on forums, etc. It's also a better long-term storage medium than home-burned CDs or DVDs.

Collapse -
what?
by froasier / January 27, 2008 12:26 AM PST
In reply to: Compression artifacts

As zip and rar are lossless algorithms themselves, they should not "corrupt" any file, regardless of previous compression. I don't know where you pulled that one from.

Collapse -
Read more carefully
by genotypewriter / January 31, 2008 5:52 PM PST
In reply to: what?

"Plus if you're zipping or rar'ing your files, you'll corrupt the files if they're not lossless."

Not lossless == lossy

Of course, if rar and zip are not lossless, you files are going to get corrupted!

Collapse -
Read more carefully
by genotypewriter / January 31, 2008 5:54 PM PST
In reply to: what?

"Plus if you're zipping or rar'ing your files, you'll corrupt the files if they're not lossless."

Not lossless == lossy

Of course, if rar and zip are not lossless, you files are going to get corrupted!

Collapse -
Lossless and lossy compression
by larrymcg / January 25, 2008 2:30 PM PST

"Lossless" comes up in the context of data compression algorithms. A lossless compression algorithm allows the original data to be reconstructed exactly. A lossy algorithm allows something close the the original data to be reconstructed.

There are several compression algorithms that we run into all the time. MP3 is a lossy algorithm used to compress sound. It tries to do it in a way that the reconstructed (or uncompressed) data sounds about the same as the original. There is a tradeoff -- as the compression is increased the resulting data (and thus file size) is reduced but the difference between the original data and the uncompressed data is greater.

Another lossy algorithm we see a lot is JPEG and GIF for photos. These algorithms try to compress the data so that the reconstructed data (i.e., photo) looks "the same" as the original. Tradeoffs similar to those of MP3 apply.

Contrary to one of the other posters, there are plenty of lossless compression algorithms. One we see all the time is a ZIP file. A ZIP file is a lot smaller than the original but the reconstructed data is EXACTLY the same as the original data. Otherwise you couldn't use it to successfully compress a computer program.

--Larry

Collapse -
GIF is lossless
by genotypewriter / January 25, 2008 10:22 PM PST

Just because GIF gives poor quality, people think it's a lossy compression format. If it was truly lossy, then everytime you save the file you should lose a bit more of data. With GIF this should not happen. GIF uses a colour palette and any "loss" is only at the initial conversion to GIF.

Collapse -
aka
by froasier / January 27, 2008 12:18 AM PST
In reply to: GIF is lossless

GIF is lossless for images that already have 256 or fewer colors. Otherwise it is lossy.

Collapse -
More on GIF and Lossy (did we need more?)
by larrymcg / January 27, 2008 12:48 AM PST
In reply to: aka

genotypewriter is being careful to talk about the compression algorithm, not the entire process of converting a photo to a GIF. The compression algorithm is lossless. However, the conversion process involves two steps: First it processes the photo w.r.t. the palette, which can result in loss of information. Then it compresses the result of the first step.

In my original post I called GIF lossy because when you start with a photo from a digital camera, and convert it to a GIF, you generally don't get back something that looks like the original. Most people would decide, based on the visual evidence, that GIF is lossy.

I looked into Wikipedia and they call GIF lossless too. But those articles are titled Lossless Compression and Lossy Compression. In that context the GIF compression algorithm is lossless.

Have we beat this to death yet?

--Larry

Collapse -
Thanks
by genotypewriter / February 2, 2008 9:23 AM PST

Yes, any form of digital encoding leads to a loss of quality that's there. That's why some people like vinyl over audio CDs when audio CDs are also uncompressed. This does not mean the audio CD or GIF is lossless... it's just the way they encode.

Collapse -
Lossless
by thanksbutnothanks / January 25, 2008 11:55 PM PST

Lossless compression is a type of compression algorithm that allows data to be compressed and then decompressed without losing any of the original data. The reason we compress data is to take a large file and convert it into a smaller file. Once it is compressed into a smaller file we can fit more of these compressed files onto a disc or a drive. We can also tranfer these compressed files over a network or the internet much faster.

Some types of compression alter the original data in such a way that some of the original data is lost. These are called 'lossy' compression formats. JPEG photos, MP3 audio files, and MPEG video files are some examples of lossy compression formats. Most lossy files can be opened in their compressed state. They normally don't have to be decompressed before being used. You can play MP3's with your favorite media player. JPEG photos will open in almost any photo editor or viewer.

Lossless compression formats such as Free Lossless Audio Codec (FLAC) and ZIP files usually have to be decompressed using a decompression utility before being opened for use. FLAC Frontend and WinZip can be used to decompress these files. Once they have been decompressed they should be exact copies of the original files.

Collapse -
what?
by froasier / January 27, 2008 12:12 AM PST
In reply to: Lossless

you don't have to decompress FLAC files before playing them. the whole point of lossless audio codecs is playability--that's what differentiates them from general compression methods like zip (that and optimization for audio files, but still).

Collapse -
Perspective
by roshanmani / January 27, 2008 2:05 PM PST
In reply to: Lossless

If the question is viewed in the perspective of audio, I've used standard rip tools that convert CD Audio into lossless formats (MP3 and WMA) that can be stored and played on a computer system. Besides the fact that these formats take a LOT of disk-space, the audio quality is as good as the original CD Audio (i.e., I haven't noticed any audible differences). I use these lossless formats only to rip and store my audio collection, from a portability perspective. So I carry my harddisk, instead of about 300 CD's.

So, are these lossless formats indeed lossless, in the perspective that the algorithms just convert from one format to another?

Collapse -
Lossless v lossy
by aflockhart / January 27, 2008 6:30 PM PST
In reply to: Perspective

Imagine your file laid out as a long stream of letters. (In fact, when you encode a file and turn it into a sequence of bytes, that's quite close to what it is.)

A LOSSLESS compression scheme would look for patterns in this long string that can be expressed in a shorter form. So instead of "xaaaaaaaabcdddebcdbcdebcd" (length 25) it might use "x8abcdddebcdbcdebcd" (19). It might also look for patterns that turn up regularly in the string and store these in a 'lookup table' so the above might become "x8aZddeZZeZ:Z=bcd" (17). And so on, with other smarter ways to shorten the string. The storage required is getting less, but you can ALWAYS decode the short string to get back to the exact original version.

A LOSSY compression scheme is based on making an judgement/decision on what can actually be perceived in the original file. In an audio file, it might choose to lose some high frequencies that are beyond the hearing range of most people.

The equivalent in our string of characters would be to decide that - for example- "most people can't detect a single letter e in a string, so if we find one we can delete it"

Our original string then becomes "xaaaaaaaabcdddbcdbcdbcd" (23) - but we can't get back from this to the original because we have thrown away the information about where the "e"s were found.

We can take this new string and compress it further using the lossless method "x8aZddZZZ:z=bcd" (15). If we repeat the process to compress the compressed string again, we can get down to "x8aZdd3Z:z=bcd" (14).

But when we reverse the process we can only get back to the 23 letter short version of the string.

Does this matter ? Depends if you can detect the fact that there is missing data. And that depends on both the nature of the data being compressed and the way it will be accessed.

For audio: can EVERY component that you are going to use to play and listen to the file - player, amplifier, loudspeakers, your ears and your brain - reproduce or detect the difference between the lossy file and the original ? If so, then it matters. But if any component in the chain can't reproduce or detect any difference, then it doesn't matter because they are going to sound the same.

Collapse -
Thanks all...
by Rollbar / January 27, 2008 11:58 PM PST
In reply to: Lossless v lossy

I appreciate all the help. I think I get the basics enough to have an idea what's going on.

aflockheart examples were really helpful too

Gary

Collapse -
(NT) Yeah..I think aflockhart's reply is a keeper!
by ahtoi / January 28, 2008 12:10 AM PST
In reply to: Thanks all...
Collapse -
Gary, didn't you get more detailed explanations than you
by NM_Bill / January 28, 2008 2:29 AM PST
In reply to: Thanks all...

probably needed?

Collapse -
LOL.... I was taking the view
by Rollbar / January 28, 2008 2:42 AM PST

that this must have been a topic a number of folks wanted to opine on and my question opened the door.

I got what I wanted, an idea of what the difference is in Lossy and Lossless.

The rest is for the techs to sort out I think.... lot of techs here in this forum though. Happy

Popular Forums
icon
Computer Newbies 10,686 discussions
icon
Computer Help 54,365 discussions
icon
Laptops 21,181 discussions
icon
Networking & Wireless 16,313 discussions
icon
Phones 17,137 discussions
icon
Security 31,287 discussions
icon
TVs & Home Theaters 22,101 discussions
icon
Windows 7 8,164 discussions
icon
Windows 10 2,657 discussions

The Samsung RF23M8090SG

One of the best French door fridges we've tested

A good-looking fridge with useful features like an auto-filling water pitcher and a temperature-adjustable "FlexZone" drawer. It was a near-flawless performer in our cooling tests.