MQN

Anything to do with computer audio, hardware, software etc.
sbgk
Posts: 1950
Joined: Mon Oct 07, 2013 9:45 pm

Re: MQN

Post by sbgk »

v24 has the fantastic detail of v23 and more solid bass, what's not to like. Nearly there.

v24 had a small error

uploaded v25 which fixes it. I would be happy to go with this one.

Some are still favouring v15 ldn, haven't listened to it for a while, anybody else like it above v25 ?
janh
Posts: 88
Joined: Wed Oct 30, 2013 9:24 pm

Re: MQN

Post by janh »

Listened to mqnplay v21, v23 and v24.
v21 still has a small but noticable veil, which I dislike. and detail is not quite as good as v23, v24, so I forget v21.

v23 and v24 are both very fine, no veil, with very good detail. differances more up to systems?
For me in orchestral pieces with mostly string sound, v23 is darker sounding, high violins less prominent than v24. best?
high voices, sopranos and tenor: I like v23 best, a little more smooth and more colour. but I only listened to one record.
piano: v23 also a little more colour than v24, maybe(?).

They are both very good, more listening needed.
Gigabyte H97M-D3H with PPA OCXO module. i7-4790T, 800MHz. 8GB Ram, 800MHz.
PPA 2-rail LPSU & Pico. JCAT battery for OS-SSD and PPA v3 USBcard.
Server 2012 R2, AO 1.40. APL HiFi DAC-S, upd. Only use 1644 .wav
sbgk
Posts: 1950
Joined: Mon Oct 07, 2013 9:45 pm

Re: MQN

Post by sbgk »

janh, don't know if you saw I just uploaded v25, think it is the most correct as far as alignment goes, amazing snap on the bass and drums, quite intense to listen to.

only has 3 instructions between the wait and the device call, so in theory should have minimum timing issues. The second syscall is aligned on a 16 byte boundary which is optimal.

syscall ;wait ; 40001095 _ 0F 05
mov rdx, r14 ; 40001097 _ 49: 8B. D6
mov r10, rdi ; 4000109A _ 4C: 8B. D7
mov rax, rsi ; 4000109D _ 48: 8B. C6
syscall ;device ; 400010A0 _ 0F 05
janh
Posts: 88
Joined: Wed Oct 30, 2013 9:24 pm

Re: MQN

Post by janh »

sbgk wrote:janh, don't know if you saw I just uploaded v25, think it is the most correct as far as alignment goes, amazing snap on the bass and drums, quite intense to listen to.
Just saw it. Tomorrow.
Gigabyte H97M-D3H with PPA OCXO module. i7-4790T, 800MHz. 8GB Ram, 800MHz.
PPA 2-rail LPSU & Pico. JCAT battery for OS-SSD and PPA v3 USBcard.
Server 2012 R2, AO 1.40. APL HiFi DAC-S, upd. Only use 1644 .wav
sbgk
Posts: 1950
Joined: Mon Oct 07, 2013 9:45 pm

Re: MQN

Post by sbgk »

think both syscalls need to be aligned, v25 a bit lacking in treble

uploaded v26 which is hopefully better.

think the alignment instructions affect the sq, so v27 just aligns the wait instruction.
TioFrancotirdor
Posts: 52
Joined: Thu Sep 18, 2014 10:29 am

Re: MQN

Post by TioFrancotirdor »

For me the latest mqnplayer v15 ldn with hardcoded device name is still the best. I am using it with newest mqncontrol and loader for haswell (best combination so far). mqnplayer with device name in text file gives a little haze in my setup.
Most preferably I would like to listen to the newest mqnplayers but with hardcoded/fixed device names ldn.
ASUS-H81i Plus, i3 4360, 8GB RAM, Linear PSU. USB/PCI PPA Studio V2.
Ubuntu Live USB in RAM
Soekris R2R Salas Ref D powered -> Modulus 86 -> MA Silver 8
sebna
Posts: 224
Joined: Mon Feb 04, 2013 9:59 pm

Re: MQN

Post by sebna »

That is interesting.

I find the mqnplay ldn15 with its original mqnloader and mqncontol (v7 both) to be best by quite a margin. Actually I think play LDN15 + control and loader 7 are nothing short of amazing and if I could only load any size test.wav with this combo I would probably live happily ever after with them :)

I have also been trying ldn15 mqnplay with new controls and loader but they are not as good as original matching, unfortunately the gap is quite big IMHO. The sound is tiring and loosing in dimensionality and presence in compare to original LDN15 and its control and loader, no matter which combination of any new loaders and control with Mqnplay ldn 15 I try (and I also find play 15 ldn better then any of the new MQNplays up to 21 where my testing has finished so far).

Still have to try anything past 21. But I can confirm that 21 was step back over 20 in clarity. 21 is hazed in compare to 20.

I cannot remember which but I narrowed it down to either MQNloader or Control in their new versions after 16 causing a lot of strain not MQNplays itself which were affecting other areas of sound but strain was mostly down to one of those two (control and loader), so probably some improvements to be had there.

Cheers
i5@800mhz haswell, 16gb @800mhz, H87 mb with PPA TCXO, PPA V1 USB
no storage of any kind, SATA disabled in BIOS, RAMos, W2012 R2, AO 1.26, MQN
Teradak ATX LPSU 210W & 5v LPSU for clean 5v to DAC
Meitner MA-1, Primare Pre30 + A33.2, Zingali HM 2.10+
sbgk
Posts: 1950
Joined: Mon Oct 07, 2013 9:45 pm

Re: MQN

Post by sbgk »

keeping the device id in the play exe will make it less user friendly, but I guess if it's the best sounding then it should be kept.

Still learning about the alignment, it's not intuitive.
TioFrancotirdor
Posts: 52
Joined: Thu Sep 18, 2014 10:29 am

Re: MQN

Post by TioFrancotirdor »

Hi Gordon,

What about this:
1. mqndeviceinfo.exe >> mqnparam.txt
2. mqnDeviceNameReplacer.exe
- reads what is in mqnparam.txt
- opens mqnplayer.exe for edit
- replaces the device name and fills with 0s at the end if needed

Both sound quality and user friendliness is saved.

Me personally I do not mind using HexEditor.
ASUS-H81i Plus, i3 4360, 8GB RAM, Linear PSU. USB/PCI PPA Studio V2.
Ubuntu Live USB in RAM
Soekris R2R Salas Ref D powered -> Modulus 86 -> MA Silver 8
grisaia
Posts: 14
Joined: Thu Dec 11, 2014 11:05 pm

Re: MQN

Post by grisaia »

TioFrancotirdor wrote:Hi Gordon,

What about this:
1. mqndeviceinfo.exe >> mqnparam.txt
2. mqnDeviceNameReplacer.exe
- reads what is in mqnparam.txt
- opens mqnplayer.exe for edit
- replaces the device name and fills with 0s at the end if needed
Sounds fine to me. But device name string has to be way longer to prevent situation in which user's device name string longer than original one
Post Reply