Tír Na HiFi

lekt wrote:

goon-heaven wrote:I start to suspect my lack of head-bang may ...

yes, goon-heaven, simple method is testing by feeling. foot tapping, head banging means good tempo, good flow,...
like your comment!

best method is vibration hehe. i am learning now. makes tweaking x100 times faster.

just listen to 2.82 again.
this version is weird.
i was sure it has very good stage depth at the first 10 seconds
after 10 seconds, things start to get 'smaller' and further back. stage depth disappears and stage becomes flat.

virus infection haha

uploaded:
v2.82.1 256 R1
v2.82.1 256 R2

R1 uses normal render loop, R2 uses render loop as v2.81.6
(i feel v2.81.6 sound good, clarity and energy. only not optimized code so its sound less air)

try these 2 versions R1 & R2, i prefer both.

lekt wrote:uploaded:
v2.82.1 256 R1
v2.82.1 256 R2

R1 uses normal render loop, R2 uses render loop as v2.81.6
(i feel v2.81.6 sound good, clarity and energy. only not optimized code so its sound less air)

try these 2 versions R1 & R2, i prefer both.

R2's bass vibration much better more refined. has pops.
R2 intense > R1.
so... R1's vocal better. R2's rock better.

2.82x's energy not as natural and free like 2.4.3.

my favorite pop version is 2.81.4 and 2.81.5. they are like 2.4.3 and with x10 vibration.

i still use s4. think it is my favorite 2 files model. it makes good sound in all situations.
actually, it makes the richest, airy, most peaceful sound with good emotion zero-intense. in my setup, it has richer sound than all software players on earth. it is a bit soft for rock. excellent for vocal.

my favorite
s4, s8
2.4.3, 2.79.2, 2.81.4, 2.81.5.

(i might have misjudged 2.81.5 earlier)

uploaded:
v2.83 256 R1
v2.83 256 R2

more clean, clarity. R2 give good guitar sound, i like it. try them.

has review v2.75.1 s4, most rich sound, thin and good depth, only bass have imbalance 0.1% but ~0% when OS is good. will back to 2 files model, now have few new methods can be good for it.

lekt wrote:uploaded:
v2.83 256 R1
v2.83 256 R2

more clean, clarity. R2 give good guitar sound, i like it. try them.

has review v2.75.1 s4, most rich sound, thin and good depth, only bass have imbalance 0.1% but ~0% when OS is good. will back to 2 files model, now have few new methods can be good for it.

all things sound great. almost the same between R1 and R2.
prefer R1

i finally understand! 2.8x is not intense. just that sound density seems high. 101% for R1. 102% for R2.
(if timerresolution=0.5ms, then sound density=103%. difficult for ears)

yes guitar sound is coming together! sharp+fast

have difference between R1 and R2, JC. R1 seems more correct but why i feel R2 give more emotion, need test more.
think almost C functions give more rich and density, cleaness also, than C++ . need improve vibration so make sound more soft, that's reduce density. most good balance is v2.4.3, always ~ 50/50, stable.

lekt wrote:have difference between R1 and R2, JC. R1 seems more correct but why i feel R2 give more emotion, need test more.
think almost C functions give more rich and density, cleaness also, than C++ . need improve vibration so make sound more soft, that's reduce density. most good balance is v2.4.3, always ~ 50/50, stable.

2.4.3 352 gives best emotion for me.
has to be magic number 352 or 320 to have complete emotion. don't know why.

lekt wrote:have difference between R1 and R2, JC. R1 seems more correct but why i feel R2 give more emotion, need test more.
think almost C functions give more rich and density, cleaness also, than C++ . need improve vibration so make sound more soft, that's reduce density. most good balance is v2.4.3, always ~ 50/50, stable.

just fixed another tweak. best to have intel hd graphic=balance power plan. same as you now, i think.

now... the story is totally changed....

density is fine for R1. seems like my new favorite.
density is still very high for R2. makes a little intense. maybe make false emotion?!

my OS is at tipping point. 1 tweak = 1-10% difference. hoping to stabilize very soon.

i see lekt.exe=71KB.

that sounds like thousands lines of code. must be very challenging to maintain it? assuming each line affects sound

i have a hard time maintaining 100-200 tweaks. all i want is every one of them to be correct and accumulative. still not there yet hehehe...