Page 261 of 299

Re: lekt player

Posted: Wed Dec 02, 2015 10:08 pm
by goon-heaven
lekt wrote:uploaded v3.47 256 act

rewriten several function of audio core, too much code and file size now bigger (4.5kb), seems negative affect to sound. in my hard disk maybe it is fragmented coz size > 4kb. i think need optimize code for reducing file size.
try it, SQ better. balance have little issue again, more code and more wrong code alignment, need some correction.
Shows much promise... more micro detail but not as clean as r12d.

Re: lekt player

Posted: Thu Dec 03, 2015 7:04 pm
by lekt
yes, v3.47 256 act sound better but defect in sentisive thing: ignored 2 settings from original code. retested, i'm wrong, need restore these settings.
how difficult with shared memory manipulation, Microsoft uses hundred settings in code. now player in exclusive mode, working by only one thread, but why if bypass critical section then SQ lesser. i can remove almost thing in first and second layer, player works in my case, but SQ lesser than.

JC, Stud_PE shows that compiled file by vs2015 tool contains debug info inside, but vs2013 not. you can see in Data_Directories\Debug. i remove this by method: in "Data Dir:" combobox select IMAGE_DIR_ENTRY_DEBUG, click GoHex, and set all nonzero bytes = 0, then click "Save to File". after that go to Data_Directories\Debug, set all nonzero bytes = 0, click "Save to File". Player file after removing debug info in PE give better sound. do this testing.

Re: lekt player

Posted: Fri Dec 04, 2015 3:52 am
by jesuscheung
yes remove debug sounds better.
better for 3.46.2 256 r12d
better for 3.47 256 act
better for 3.46.3 256 srv

Re: lekt player

Posted: Fri Dec 04, 2015 3:53 am
by jesuscheung
goon-heaven wrote:
lekt wrote:uploaded v3.47 256 act

rewriten several function of audio core, too much code and file size now bigger (4.5kb), seems negative affect to sound. in my hard disk maybe it is fragmented coz size > 4kb. i think need optimize code for reducing file size.
try it, SQ better. balance have little issue again, more code and more wrong code alignment, need some correction.
Shows much promise... more micro detail but not as clean as r12d.
yea, am back to srv, and r12d.

Re: lekt player

Posted: Fri Dec 04, 2015 4:08 am
by jesuscheung
majorlinkversion=E now that vs2015 is used as the compiler.

not sure. A, B, C, E all sound different. seems E is best

Re: lekt player

Posted: Fri Dec 04, 2015 5:49 am
by lekt
jesuscheung wrote:majorlinkversion=E now that vs2015 is used as the compiler.

not sure. A, B, C, E all sound different. seems E is best
i've also detected this, tried all. E better.

Re: lekt player

Posted: Sat Dec 05, 2015 11:42 pm
by lekt
uploaded:
v3.48 256 srw
v3.48.1 256

file size 4.5kb, still can't reduce size coz restored some original settings. need more optimize.
both versions are good, sound more refined.
v3.46.2 256 r12d have good setting, will learn this more and apply for v3.48x later.

Re: lekt player

Posted: Sun Dec 06, 2015 10:04 pm
by goon-heaven
Yes, both these 48s are very good, with 48.1 having the better SQ here and closer to r12d...

But I still prefer to listen through r12d with music that I often hear live. What is this good setting of r12d?

I wait for your v3.48.2 256 r12d...

Re: lekt player

Posted: Mon Dec 07, 2015 8:02 am
by jesuscheung
lekt wrote:uploaded:
v3.48 256 srw
v3.48.1 256

file size 4.5kb, still can't reduce size coz restored some original settings. need more optimize.
both versions are good, sound more refined.
v3.46.2 256 r12d have good setting, will learn this more and apply for v3.48x later.
still prefer srv, r12d.

3.48.1 maybe better 3.48 256 srw.
both versions vocal lesser than srv.

Re: lekt player

Posted: Mon Dec 07, 2015 9:36 pm
by lekt
uploaded:
v3.48.2 256 r12d
v3.48.3 256
v3.48.4 256
v3.49 256 opt

v3.48.2 and v3.48.3 used setting of v3.46.2 256 r12d but now file size bigger, different sound. think need goto small size, no bigger 4kb, long code not good, dificult make code alignment.
prefer v3.49 256 opt, think it more clean. try compare, here's micro difference.