Tír Na HiFi

Octagon wrote:

sbgk wrote:Am also considering a wav file program to read in and optimally write an optimised wav file...

Is that the reason why 3.70 takes some seconds to start? Are you reorganizing the way of writing the music file into RAM?

no, it was missing the start of the track so added some seconds to avoid that. Have changed mqnplay to start playing as soon as it gets some data, but that's only on my own version so far.

lukivision wrote: Is there a way, to use MQN as foobar´s sound engine?
Luki

Are you referring to maintain foobar as gui and use mqn as player?

sbgk wrote: am further optimising control. Am also considering a wav file program to read in and optimally write an optimised wav file so it can be demonstrated on other players. Wonder if it would make a difference to other types of files eg jpg or raw files.

Interesting investigation - please post your results when you have done this :)

m.massimo wrote:

lukivision wrote: Is there a way, to use MQN as foobar´s sound engine?
Luki

Are you referring to maintain foobar as gui and use mqn as player?

Yes, indeed. If possible, I would prefer to use just foobar with MQN fully integrated.

Luki

Yes it's possible. First of all, download MQnfoobar.txt and look at it carefully.

I use foobar as gui and mqn as player. I use mqnload too cause my library is in .flac and it needs to be decoded to .wav before playing.
I use foobar along with mqn in a track by track way cause I want foobar to show the tags of every single track.
With this method you have control of any single track, but if you have a .flac library you need to introduce a delay between two consecutive tracks to allow the decoding process to be concluded otherwise when foobar launches the next track you've the final seconds of the actual one truncated. The actual delay depends on the file size, so if you play SD and HD mixed, you must take into consideration the most lengthy decoding time and increase the delay accordingly. For a 1644 it's a matter of 6/7 seconds or so to have the sound starting (with 3.64, with 3.70 this time must be more than doubled). For live recordings this is a real problem and in that case I select a group of tracks inside foobar (you can do in any file manager as well), copy and then launch mqnload.exe that decodes the entire batch and then play it. The advantage in this case is that you don't have delays between tracks anymore, but you lose control of every single track.

Let me know if you need more specific info on my setup.

about FLAC 1.3.1 performance

great performance improvement

1.3.1 vs 1.3.0 time needed flac <- -> wav
Level 5

1644 on BayTrail J1900 Win7-32
Encoding -35%
Decoding -12%

2496 on BayTrail J1900 Win7-32
Encoding -36%
Decoding -45%

1644 on Atom AD2700 Win8.1-64
Encoding -25%
Decoding -24%

2496 on Atom AD2700 Win8.1-64
Encoding -14%
Decoding -20%

m.massimo wrote:Yes it's possible. First of all, download MQnfoobar.txt and look at it carefully.
...
Let me know if you need more specific info on my setup.

Thank you, I'll try! But it doesn't seem to be real comfortable, rather compromising.

Luki

Octagon wrote:3.70 offers great sq. As mentioned the glitch is away and also the growing RAM amount of play.exe is stopped after playing.
I recognize a gap of 15-18 seconds before playing starts. (edit)

The performance with OS in RAM is incredible, with MQn CPU usage is below 1% including online convolving via VSTHost. Compared to JPlay that is far less allthough JPlay is pretty well with 5% CPU usage including online convolving with foobar's fooconvolver. If one is interested I might post some screenshots from taskmanager.

3.70 fixes the glitch and sounds really good.

However on my rig, the RAM usage going at 200-300KB per sec does still continue at the end of the track(s) until the 'X' is pressed.

uploaded 3.72, a setting wasn't correct.

3.72 is excellent but doesn't play 24/96 just a blast of noise, great for getting me out of the chair.
KM