Newbie here. Video Compression Query

[gargz]

I need to compress my mp4 and avi files into a zip/rar or any other format for storage purposes only and I don't want any quality loss.
For example, I downloaded a file where videos were compressed to less than half their original size but after an hour or two of extracting, I got the full sized videos.

Any help would be appreciated guys.

P.S. Don't have the option of buying new HDD.

[hech54]

I need to compress my mp4 and avi files into a zip/rar or any other format for storage purposes only and I don't want any quality loss.
For example, I downloaded a file where videos were compressed to less than half their original size but after an hour or two of extracting, I got the full sized videos.

Any help would be appreciated guys.

P.S. Don't have the option of buying new HDD.

There is no quality loss with Zip/Rar. Half the size of the original?.....you are mistaken.

[DB83]

You may get a 2 or 3% size reduction with a zip/rar file. But that is not the purpose of such files.

What you are seeing in your 50% is just one rar/zip part whereas the whole file is contained in more than one - it is done that way to enabled the parts to be hosted.

The only way to actually make the video smaller is to reduce the bitrate and then you WILL lose quality.

[Cornucopia]

Since most AVC-in-MP4 files are already ~1/300th of their original uncompressed size, do you honestly think zip or rar or 7z could find remaining redundant data to give you 1/600th?

First, the algorithms used in zip/rar/7z are ALSO already being used in AVC, etc.
Second, there is already nearly complete entropy in an AVC-type file, so it is a very good possibility that not only would you not get 50% reduction, but you wouldn't even get the 2 or 3% reduction - you might actually INCREASE your filesize.

You better rethink that HDD option or accept the likelihood of quality loss (via bitrate-reduced re-encoding such as DB83 mentioned). Or make use of the cloud.

Scott

[gargz]

Check This Torrent
https://kickass.so/iris-reading-speed-reading-foundation-course-t9275162.html
Its actual size is 5 GB

and now check this
http://piratebb.org/Thread-Iris-Reading-Speed-Reading-Foundation-Course?highlight=iris+reading
Size is 2.91 GB(After Extraction: 5GB)

Hows that possible.

[poisondeathray]

If there is a substantial reduction in filesize, it means something is wrong, or original wasn't compressed very well (not efficient audio or video compression to begin with)

If you have highly compressed video, using winrar , winzip can actually INCREASE the filesize

So it really depends on what you start with

[gargz]

Can somebody download just one of the mp4 files from the link and confirm that it is due to inefficient coding. That'd be very helpful. Or tell me how to check if it is efficient or not.

[jagabo]

It doesn't matter what's going on in that particular torrent. The last step of all high compression codecs is entropy encoding, just like that used in WinRar, ZIP, etc. So A/V files encoded with those codecs won't be much smaller after archiving.

[poisondeathray]

Just take one of your random files and use winrar or some archiving software. Test it out yourself.

The average compression gain might be <1% . On some files , highly compressed files, filesize actually gets larger. Lower compressed files like DV with uncompressed audio might gain 10% reduction in filesize, because they weren't as heavily compressed to begin with

The other reason there might be a larger than 1% size discrepancy, is people use recovery records as another method of security. So you might add a 10% recovery record for the archive for example (in case archive is damaged, you can ofter recover the contents)

The other possiblity is the archive is damaged or incomplete (video files aren't intact)

[DB83]

And you expect someone to download a 3 gb file just to prove a point ?

For all we know that tutorial could be a series of still images and audio. That may compress better.

Someone also needs to aquaint themselves with forum rules regards to warez (even tho that may be an indirect consequence in this topic)

[NoBuddy]

Just did a test on a highly compressed mp4 file (mostly static images, very little moves, instructional)
Even so getting more than 10% reduction on this file is quite good, getting anything near 50% would be imposible for this file.
Used programs 7-zip, WinRar, WinUHA-UHARC - always with Best Compression activated.

File-Info:
General
Complete name : D:\Chord Patterns in 'F' Major.mp4
Format : MPEG-4
Format profile : Base Media
Codec ID : isom
File size : 8.86 MiB
Duration : 2mn 55s
Overall bit rate mode : Variable
Overall bit rate : 423 Kbps
Encoded date : UTC 2014-10-01 17:19:04
Tagged date : UTC 2014-10-01 17:19:04

Video
ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : Baseline@L3.0
Format settings, CABAC : No
Format settings, ReFrames : 3 frames
Format settings, GOP : M=1, N=75
Codec ID : avc1
Codec ID/Info : Advanced Video Coding
Duration : 2mn 55s
Bit rate : 323 Kbps
Maximum bit rate : 696 Kbps
Width : 640 pixels
Height : 360 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 25.000 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.056
Stream size : 6.77 MiB (76%)
Writing library : Zencoder Video Encoding System
Encoded date : UTC 2014-10-01 17:18:53
Tagged date : UTC 2014-10-01 17:19:04

Audio
ID : 2
Format : AAC
Format/Info : Advanced Audio Codec
Format profile : LC
Codec ID : 40
Duration : 2mn 55s
Bit rate mode : Variable
Bit rate : 96.0 Kbps
Maximum bit rate : 107 Kbps
Channel(s) : 2 channels
Channel positions : Front: L R
Sampling rate : 44.1 KHz
Compression mode : Lossy
Stream size : 2.04 MiB (23%)
Encoded date : UTC 2014-10-01 17:19:04
Tagged date : UTC 2014-10-01 17:19:04

[poisondeathray]

Just did a test on a highly compressed mp4 file (mostly static images, very little moves, instructional)
Even so getting more than 10% reduction on this file is quite good, getting anything near 50% would be imposible for this file.
Used programs 7-zip, WinRar, WinUHA-UHARC - always with Best Compression activated.

Format profile : Baseline@L3.0
Format settings, CABAC : No

I wouldn't quite call it "highly" compressed. You can tell from mediainfo that the video is not optimally compressed - it uses baseline profile (no b-frames, no cabac) . So typical compression gains you would expect will be larger.

Take a more commonly used compression scheme, e.g. x264 with high profile (uses cabac and b-frames), and your typical compression gains would be <1% . You tend to gain a bit more with audio than highly compressed video like x264. Less compressed video like xvid would typically gain 1-2% in compression

eg. video only, using x264 (no audio) . They all are larger, except for uharc, which is 0.05% smaller . If anyone wants I can post link to video

original 10,787,390 bytes

winrar default 10,787,480 bytes

7zip default 10,927,963 bytes

uharc default 10,782,399 bytes

[newpball]

Can somebody download just one of the mp4 files from the link and confirm that it is due to inefficient coding. That'd be very helpful. Or tell me how to check if it is efficient or not.

I'd say NO!

People here give sane advice, compressed video is already eh, well you guessed, compressed. Compressing something that is already compressed is futile.

[jagabo]

I downloaded one MP4 file from the torrent: "Foundation Course Part 1.mp4", 537,316,962 bytes. Using Windows built in ZIP compressor reduced it to a 534,893,567 byte ZIP file. Not even half a percent smaller.

[NoBuddy]

poisondeathray is right
My compressed, small mp4 file was made for a weird mobile device, low bitrate to get it under 10Mb - we had to send it by email.
Now I used High profile, and WinRar doesn´t even reach a 1% gain.
So, as all and everybody already told you, don´t expect magic.

Format profile : High@L4.1
Format settings, CABAC : Yes
Format settings, ReFrames : 6 frames
Writing library : x264 core 130 r2273 b3065e6
Encoding settings : cabac=1 / ref=6 / deblock=1:1:1 / analyse=0x3:0x113 / me=hex / subme=7 / psy=1 / psy_rd=0.40:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 / trellis=1 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 / threads=1 / lookahead_threads=1 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 / constrained_intra=0 / bframes=5 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=1 / weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc_lookahead=40 / rc=abr / mbtree=1 / bitrate=300 / ratetol=1.0 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / vbv_maxrate=62500 / vbv_bufsize=78125 / nal_hrd=none / ip_ratio=1.40 / aq=1:0.60

[poisondeathray]

I downloaded one MP4 file from the torrent: "Foundation Course Part 1.mp4", 537,316,962 bytes. Using Windows built in ZIP compressor reduced it to a 534,893,567 byte ZIP file. Not even half a percent smaller.

This observation would suggest the smaller version was re-encoded, and not really the same quality

[gargz]

I downloaded one MP4 file from the torrent: "Foundation Course Part 1.mp4", 537,316,962 bytes. Using Windows built in ZIP compressor reduced it to a 534,893,567 byte ZIP file. Not even half a percent smaller.

This observation would suggest the smaller version was re-encoded, and not really the same quality

I can confirm that the smaller version given in the second link is absolutely the same quality as the original. Only difference is that we are not able to play the smaller version directly as it is WinRAR file and after decompression/extraction of the WinRAR file, we get the original 5.07 GB file. All the mp4 files are full fledged videos and not still images.

P.S. I have heard there are softwares that can reduce a file to 1% of its size but take days to Compress/Decompress. Maybe it has to do something with it.

[poisondeathray]

I downloaded one MP4 file from the torrent: "Foundation Course Part 1.mp4", 537,316,962 bytes. Using Windows built in ZIP compressor reduced it to a 534,893,567 byte ZIP file. Not even half a percent smaller.

This observation would suggest the smaller version was re-encoded, and not really the same quality

I can confirm that the smaller version given in the second link is absolutely the same quality as the original. Only difference is that we are not able to play the smaller version directly as it is WinRAR file and after decompression/extraction of the WinRAR file, we get the original 5.07 GB file. All the mp4 files are full fledged videos and not still images.

P.S. I have heard there are softwares that can reduce a file to 1% of its size but take days to Compress/Decompress. Maybe it has to do something with it.

And how did you "confirm" it ?

If jagabo's zip test was <1% difference, and was representative of the rest of the files, the most you will get with any archiving software is a few % at most

Yes, you can re-encode it with some loss (visually looks very close) to about 1/2 the size. But this isn't lossless, and not archival

You can play video files in winrar archives directly with some media players . (Maybe not the newer Rar5 format)



What softwares can reduce an already compressed file to 1% of the filesize losslessly ? Because if it existed, the author could sell it for billions of dollars (with a "B" )

[gargz]

I downloaded one MP4 file from the torrent: "Foundation Course Part 1.mp4", 537,316,962 bytes. Using Windows built in ZIP compressor reduced it to a 534,893,567 byte ZIP file. Not even half a percent smaller.

This observation would suggest the smaller version was re-encoded, and not really the same quality

I can confirm that the smaller version given in the second link is absolutely the same quality as the original. Only difference is that we are not able to play the smaller version directly as it is WinRAR file and after decompression/extraction of the WinRAR file, we get the original 5.07 GB file. All the mp4 files are full fledged videos and not still images.

P.S. I have heard there are softwares that can reduce a file to 1% of its size but take days to Compress/Decompress. Maybe it has to do something with it.

And how did you "confirm" it ?

If jagabo's zip test was <1% difference, and was representative of the rest of the files, the most you will get with any archiving software is a few % at most

Yes, you can re-encode it with some loss (visually looks very close) to about 1/2 the size. But this isn't lossless, and not archival

You can play video files in winrar archives directly with some media players . (Maybe not the newer Rar5 format)



What softwares can reduce an already compressed file to 1% of the filesize losslessly ? Because if it existed, the author could sell it for billions of dollars (with a "B" )

I can confirm because I downloaded both of the files.

[poisondeathray]

I can confirm because I downloaded both of the files.

How does "downloading" them confirm anything ? The unarchived filesize doesn't necessarily mean they are the same.

I want to know - how did you measure the video quality or file integrity ? MD5 checksum ? SSIM ? PSNR ?

If it is verified to be bit for bit identical, then I have no idea what is going on... and there is no way I'm going to download them




But you should know that even if this is bit for bit identical - these are not representative of the typical results that you will get for your files

eg. Take one of your own h264/aac/mp4 videos and use winrar, any settings, and they typically might compress about 1% or less. Try it out right now.

[newpball]

Seems to me all the advice to the poster falls on deaf ears.

[hech54]

And why is it always guys from India? I swear we've had this exact same type of discussion twice before, both India. Do they charge per-megabyte for internet in India or something? Holy Shit.

[gargz]

I can confirm because I downloaded both of the files.

How does "downloading" them confirm anything ? The unarchived filesize doesn't necessarily mean they are the same.

I want to know - how did you measure the video quality or file integrity ? MD5 checksum ? SSIM ? PSNR ?

If it is verified to be bit for bit identical, then I have no idea what is going on... and there is no way I'm going to download them




But you should know that even if this is bit for bit identical - these are not representative of the typical results that you will get for your files

eg. Take one of your own h264/aac/mp4 videos and use winrar, any settings, and they typically might compress about 1% or less. Try it out right now.

That's good. I thought if the file sizes of both the original and unarchived file were the same, they must be identical.
I have tried compressing video files using winrar before, but get the same 1% result, thats why I created the thread.

[gargz]

And why is it always guys from India? I swear we've had this exact same type of discussion twice before, both India. Do they charge per-megabyte for internet in India or something? Holy Shit.

Most Internet connections in india are data limited but I am fortunate to have an unlimited data connection, which I think is part of the problem because I download so much that now I don't have enough space to store data. I just want a way to archive data in my limited data storage.

[KBeee]

If "Zipping" a video file could reduce the size by 50%, think what that would imply.
Stick a hardware decoder chip inside a BluRay player, and suddenly your 50GB disk is 100 GB, DVD's go from 8 GB to 16 GB etc.
If it worked, don't you think manufacturers would have done this?

[jagabo]

If "Zipping" a video file could reduce the size by 50%, think what that would imply.
Stick a hardware decoder chip inside a BluRay player, and suddenly your 50GB disk is 100 GB, DVD's go from 8 GB to 16 GB etc.
If it worked, don't you think manufacturers would have done this?

Essentially, this is already the case. The video isn't in a ZIP container but it's entropy encoded just like data in a ZIP file.

[KBeee]

I think you're misreading my post.
IF just "Zipping" the finished already compressed video were possible for a 50% video filesize reduction - It would have already been done, and saved millions instead of creating new formats. Heck, you could Zip the Zip file and get it down to 25%...