Tír Na HiFi

sbgk wrote:removing getbuffer and release would be good, do you replace them with something else ? I found some code that someone had written to replace these on osx so it could run windows, can send a link if I can find it.

do you massage the data ? Is that what the fibonacci comment is about ?

the player has no file read code, the control starts the player, the player allocates memory and the control gets the address of the memory then loads it with data and then hands off to the player. communication by reading bytes in the allocated memory ie when =1 then the data is loaded and the playback can start. it's c++, but with minimised winmaincrtstartup and no dll linking. when I had the file code in the player I did convert it to c functions, but never compiled the whole thing as c, would be interested to know how to do that.

- each iteration in render loop wasapi uses one buffer (ie. 256frames/1024bytes). first buffer pointer 0xXXXXXXXX, second buffer will have pointer 0xXXXXXXXX + 1024, and thirty will back to first 0xXXXXXXXX (= seconds - 1024). so then for what we need GetBuffer? need only get it once before render loop. so then can remove releasebuffer also, player run. OK, replace GetBuffer with calculation pointer for each iteration (good address can be calculate by xor 1024). try this trick, must be much better. it is my own idea, not reading on internet.

- and interesting thing: result of h=CreateEventA(0,0,0,0) give constant interger, in my case h=300, sometime 304, always, even after restarting windows. tried with constant, of course big benefit coz it used in reder loop.

- i tried with 2 files/2 processes model, main detach child, but still affect SQ. each thing in read file process affect render process although first already terminated. so then i think need 3-files/3-process, will try. maybe can absolutely isolate read file task. I/O device very bad for audio.

- sbgk, last lekt versions doesn't use any function from msvcr120.dll but i can't compile without it. how do you make this? i see mqn works without it.

even since you told me the 3ms clockrate....
i start to like 160/320/640!

160 has flow/soundstage that is only possible with 160/320/640
448 is almost like 160 except for some lost in details

in fact, 160 is better everything

difficult to compare
2.96.2 256 nos MOD
with
2.96.4 160 c2 MOD

coz different buffer.

these two are best overall sound in my setup.

i still think nos has better natural gravity in weight

try mqn5.64 sse... the vibration density is the highest -> solid sound. it's a shame that the affinity tweak in mqn f**ks up the micro. texture quality is limited. texture has resistance. little bit earaches on headphones.

maybe you know why CD red book standard is 44100Hz, not 48000. i don't know this secret, this history. now 44100 anywhere in internet, in our life. seems people like it more than others, seems it have better sound wave vibration. DVD, DVDA: 48000/96000. CD, SACD: 44100/88200 and have more likeable than. music better.
why, why...

jesuscheung wrote:...difficult to compare
2.96.2 256 nos MOD
with
2.96.4 160 c2 MOD

coz different buffer.

these two are best overall sound in my setup...

i like micro vibration of nos.

uploaded:
v2.96.4 160 v1
v2.96.4 160 v2
v2.96.4 160 v3

3 vibration levels, v3 have most much vibration of lekt. think v3 have best SQ, but why i feel v1 very good emotion, music have sadness.

all 2.96.4xx have same code, only compiled by different methods. use new compiling method, JC, very good, although my OS now dirty but compiler can ignore much of this dirty. of course if reinstall windows then shall be much better. try and enjoy!!

lekt wrote: 3 vibration levels, v3 have most much vibration of lekt. think v3 have best SQ, but why i feel v1 very good emotion, music have sadness.

yes.

lekt wrote: all 2.96.4xx have same code, only compiled by different methods. use new compiling method, JC, very good, although my OS now dirty but compiler can ignore much of this dirty. of course if reinstall windows then shall be much better. try and enjoy!!

v1 vibration is very clearly
v2 density feels higher but vibration is less clearly
(v1+v2)/2?

don't like v3

v3 earache. i tried Stud_PE setting and this earache ignored, balance/SQ much better:
- Headers/Characteristics/Relocations stripped: check it
- Headers/DllCharacteristics/Dynamic Base: uncheck it
do you tried this MOD? try it.
for next versions i will include /DYNAMICBASE:NO into compilers commandline. do this by compiler can remove one section named .reloc in .exe file, file size reduced. but why can not remove this section by Stud_PE ?

lekt wrote:v3 earache. i tried Stud_PE setting and this earache ignored, balance/SQ much better:
- Headers/Characteristics/Relocations stripped: check it
- Headers/DllCharacteristics/Dynamic Base: uncheck it
do you tried this MOD? try it.
for next versions i will include /DYNAMICBASE:NO into compilers commandline. do this by compiler can remove one section named .reloc in .exe file, file size reduced. but why can not remove this section by Stud_PE ?

don't use stud_PE to MOD. SQ WILL DECREASE
because stud_PE also changes other things in exe

use ultraedit

few months ago, i tried DYNAMICBASE, it didn't seem to improve SQ..
maybe better for you to do by compiler?!

UPDATE
misjudged. v2 is not good. v1 is the best! v1 has beautiful vibration

uploaded mods
2.96.4 160 v1 MOD
2.96.4 160 v3 MOD

you right! v3 MOD is goooood

is it better than v1?
don't feel anything with v1 MOD

v1 vs v3 MOD
.... need time to test

UPDATE
after MOD, v3 seems to be much better tighter bass.
v3 MOD bass > v1
emotion automatically improves

i actually quite like this 2.96.4 160 v3 MOD

it vibrates like 2.90.4
it has vibration quality like 2.90.9

slightly prefer 2.90.4 vibration. more naturally
2.96.4 more refine > 2.90.4 (2.90.4 slightly raw earaches)

(MOD makes vibration less natural?)

tweaking windows is hard work....
10 steps forward
9 steps backward... kill me now...

i am almost finish.
99.5% of sound vibrates
0.5% doesn't vibrate. why

if 0.5% also vibrates, it will be entry level hifi.

0.5% is BIG issue.
it means
mqn=harsh
lekt=hard
xa=dirty+hard

all because 0.5% doesn't vibrate