MQN

[sbgk]

uploaded play v55, to see if you agree it's an improvement. still one track.

uploaded control v28, took out a step in the load which I don't think is necessary, so should load quicker.

V55 has a phenomenal transparency indeed.

what do you think of v60 ? think that's the final version.

[Aleg]

uploaded play v55, to see if you agree it's an improvement. still one track.

uploaded control v28, took out a step in the load which I don't think is necessary, so should load quicker.

V55 has a phenomenal transparency indeed.

what do you think of v60 ? think that's the final version.

Did not yet have time to download and install, will have to wait till tonight.

[Aleg]

V60 has less of the perceived transparency than v55.

But v55 can be quite sharp in its treble.

I think v55 heightens the treble thereby aiding the mid-range in displaying better the high-frequency harmonics that accompany that mid giving a very good sense of depth and transparency. It does increase the regular treble notes as well though, which can become too sharp.

I would not be able to listen to v55 for a long time, except to be wow-ed for its effect.

Would in the end go for something like V60 rather than V55.

[sbgk]

ok, after more playing around think I've got the ultimate version, even more transparency, good bass, treble.

uploaded v61.

having great fun testing it with 1980's Van Halen, which can sound pretty flat on lesser players, but very good with v61.

v62 has the best growling bass, might be the final one, unless the bass is too much.

Difficult to judge bass, live concerts seem to have plenty in comparison to what a digital hifi outputs, turntables produce great bass. We shall see.

v63 is even better.

as usual it takes a few iterations to understand what's going on and arrive at the best solution.

v65 is the real deal, shall stop there.

[Aleg]

...

v65 is the real deal, shall stop there.

V65 is a very good version indeed.
Looking forward to a multi-track loader now (so I can stop appending all my tracks into a single one when I want to listen to an album, doubles my storage space :-)

[lukivision]

I still maintain the best you can get is going to be a linux os with optimised kernel that reads the data from ram without any user app interference, this will require the driver to be modified as well. Not a small task, why has no one tried it ? Are they all deaf ?

Hi,

here is a guy who did something else:

http://frank_l.bitbucket.org/stereoutils/player.html

Luki

[TioFrancotirdor]

ok, after more playing around think I've got the ultimate version, even more transparency, good bass, treble.

uploaded v61.

having great fun testing it with 1980's Van Halen, which can sound pretty flat on lesser players, but very good with v61.

v62 has the best growling bass, might be the final one, unless the bass is too much.

Difficult to judge bass, live concerts seem to have plenty in comparison to what a digital hifi outputs, turntables produce great bass. We shall see.

v63 is even better.

as usual it takes a few iterations to understand what's going on and arrive at the best solution.

v65 is the real deal, shall stop there.

v65 is absolutely fenomenal! Thanks Grodon.

[janh]

v65 is absolutely fenomenal! Thanks Gordon.

+1

[sbgk]

v65 is absolutely fenomenal! Thanks Gordon.

+1

v66 just a little bit betterer

If someone can give me a preference between v66 and v67, would be much obliged, am using v67, but don't let that influence things.

[sbgk]

I still maintain the best you can get is going to be a linux os with optimised kernel that reads the data from ram without any user app interference, this will require the driver to be modified as well. Not a small task, why has no one tried it ? Are they all deaf ?

Hi,

here is a guy who did something else:

http://frank_l.bitbucket.org/stereoutils/player.html

Luki

doesn't look very promising, using timers and sleep (been there, done that), why do you think we went to event managed buffer fill.
aplay uses alsa.

MQn loads the data into ram using the loader, so no need for extra step he describes of loading it to ram to refresh the data.

what I was getting at was the player integrated into the kernel and driver so there is no user mode/kernel mode interaction which has a big affect, I believe (the kernel tests the buffer passed from user mode because it can't be trusted, every time the buffer is filled, which is not optimal). this would be impossible in windoze, hence the interest in linux, problem is it would then be open source.

event buffer fill is the best otherwise the code needs work out where the pointer is in the buffer and how much to fill and then handle overruns/underruns, crazy, designed by bits are bits people, no doubt.