Paq8sk

[madserb]

SSE2 and AVX2 builds here:
paq8sk43_sse2.exe
​paq8sk43_avx.exe
​

[CompressMaster]

@suryakandau
what about optimizing paq8sk for file A10.jpg from Maximum Compression Corpus? I am interested whats the limit in compression ratio.

[suryakandau@yahoo.co.id]

Yes I am interested in the limit of jpeg compression ratio. Btw how much jpeg can be compressed ?

[kaitz]

I think you can get a10 below 618xxx in 100sec or less.
619xxx in 50 sec.

[suryakandau@yahoo.co.id]

There is a trade off between compression ratio n speed... Like cmix...

[suryakandau@yahoo.co.id]

Paq8sk44
- improve jpeg compression
using -s8 option on
f.jpg (darek corpus)
Total 112038 bytes compressed to 80194 bytes.
Time 19.17 sec, used 2444 MB (2563212985 bytes) of memory
here is source code and binary file


paq8sk44 -s8 a10.jpg 842468 616629 151.72 sec 2444 MB
paq8pxd_v95 -s8 a10.jpg 842468 618555 43.42 sec 1984 MB
paq8px_v200 -8 a10.jpg 842468 624597 26.51 sec 2602 MB

[CompressMaster]

@suryakandau
Are you able to go below 550,000 bytes of A10.jpg compression? And with F.JPG below 40,000 bytes?
I am just interested if paq8sk can be tuned to achieve that ratio - regardless time and mem. usage.

btw, you have irrelevant space between attached file and text.

[suryakandau@yahoo.co.id]

I think it's difficult to go below 550000 bytes of A10 jpeg file..

[moisesmcardona]

Hey Surya, I noticed some of your releases are compiled with MT=ON and some with MT=OFF. Can you please decide to stick with either one? The command line breaks when MT=OFF. Thanks!

[suryakandau@yahoo.co.id]

if the command line breaks when MT=OFF so use MT=ON

[suryakandau@yahoo.co.id]

Paq8sk45
- improve jpeg compression
- now can use lstm model using -x option
using -s8 option on
f.jpg (darek corpus)
Total 112038 bytes compressed to 80131 bytes.
Time 21.01 sec, used 2180 MB (2285897857 bytes) of memory
a10.jpg (maximum compression corpus)
Total 842468 bytes compressed to 616305 bytes.
Time 160.95 sec, used 2180 MB (2285897891 bytes) of memory
here is the source and the binary inside the package

[CompressMaster]

@suryakandau
Very good improvement!
As for F.JPG, only 131 bytes left to break the 80 000 barrier! Good!
Your paq version is currently the BEST as for JPG compression ratio! Keep working on it!
btw, are you interested only in JPG compression improvements?

[suryakandau@yahoo.co.id]

until now i am interested only in jpeg compression improvement..

[suryakandau@yahoo.co.id]

Paq8sk46
- improve jpeg compression
using -s8 option on
f.jpg (darek corpus)
Total 112038 bytes compressed to 80087 bytes.
Time 26.14 sec, used 3285 MB (3444911171 bytes) of memory
a10.jpg (maximum compression corpus)
Total 842468 bytes compressed to 616176 bytes.
Time 192.61 sec, used 3287 MB (3446746213 bytes) of memory
there is source code and binary inside the package file
​

[suryakandau@yahoo.co.id]

Paq8sk48
- improve the compression ratio for each data type by adding some mixer context set in each predictor(DEFAULT, JPEG, EXE, image 24 bit, image 8 bit, DECA, textwrt).
- improve image 8 bit model
c.tif (darek corpus)
paq8sk48 -s8
Total 896094 bytes compressed to 270510 bytes.
Time 166.60 sec, used 2317 MB (2430078529 bytes) of memory
paq8px201 -8
271186 bytes
Time 76.12 sec, used 2479 MB (2599526258 bytes) of memory
paq8pxd101 -s8
Total 896094 bytes compressed to 271564 bytes.
Time 64.56 sec, used 2080 MB (2181876707 bytes) of memory

[Darek]

@suryakandau - which version is an base of paq8sk48 - paq8px or paq8pxd?
OK, i Know. paq8pxd is the base

[suryakandau@yahoo.co.id]

The base of paq8sk is paq8pxd. @darek, could you test it on Silesia using the same parameter like you do when you test paq8pxd ??

[suryakandau@yahoo.co.id]

I just curious how much Silesia can be compressed using the new mixer context set in each predictor...

[Darek]

OK, I'll try. Somehow I have some issues with my laptop recently - I've started 5 times enwik9 for paq8px v201 - and even after 3-4 days my computer crashes...