IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
sound quality, vst, external sound board or dac, [moved from Site Related Discussion/TOS #6]
bud eno
post Jun 20 2013, 14:17
Post #1





Group: Members
Posts: 1
Joined: 20-June 13
Member No.: 108742



hello, my name is corrado.
i'm about to update my sound system, and if possible i would like to hear some opinion about my possibilities: i'll explain.
i have two goals, one is to listen to classical music with 150-200 € headphones, being able to fully use their finesse and quality;
the other goal is to play a mute keyboard with a 30gb piano library without feeling any latency and with perfect sound.
my actual setup consist in a laptop with 2,1ghz and 2gb ram with windows xp, and a m-audio audiophile usb, very old model. the integrated audio of my 250€, bought on 2007 laptop is not bad, at least not for me that i' m used to it, and at least i won' t hear any popping noise as when using the audiophile usb both for listening and for playing.
i've been reading more and more (because i search more) about DACs, and their quality improvements, so the questions are these, do you think is possible to find an external audio card that will be good enough to exploit the full potential of my new headphones and allow me to play flawlessly heavy vst instrument on the keyboard? will i need both a DAC and an external audio card? and wich price range would be an external audio card that will allow me to fulfill both my requirements?
thanks a lot in advance, hope the question makes sense and is pertinent to this forum because i' m a real newbie, and also i sense that this might not be the right forum for these kind of questions, so sorry in advance for wasting your time.

ps. i read now i posted in site related discussions, sorry i'll wait for a moderator to move it to the proper area if there is

This post has been edited by bud eno: Jun 20 2013, 14:19
Go to the top of the page
+Quote Post
Melomane
post Jun 24 2013, 18:00
Post #2





Group: Members
Posts: 643
Joined: 13-January 03
From: Marseille
Member No.: 4552



hi, if you don't hear pop, click or noise,
and if latency is correct fot play instrument
it is not necessary to change anything.

m-audio audiophile usb is very old model, but it is very good sound card.
DACs quality improvements is snake oil.

don't worry be happy.
sorry for my not good english.



--------------------
Music is my first love.
Go to the top of the page
+Quote Post
DVDdoug
post Jun 24 2013, 20:06
Post #3





Group: Members
Posts: 2598
Joined: 24-August 07
From: Silicon Valley
Member No.: 46454



QUOTE
the other goal is to play a mute keyboard with a 30gb piano library without feeling any latency and with perfect sound.
If you are not hearing any defects (such as noise) a "better" soundcard/interface will probably not make an improvement. If you were recording audio (instead of capturing MIDI), I'd recommend an external interface, but for playback, a regular soundcard is often OK.

Wth a normal soundcard, sometimes there can be impedance-related frequency-response variations from tne line-level outputs, or you may not get enough volume. In that case a headphone amplifier may help. (You don't need a super-expensive headphone amp, just something that's designed to drive headphones.)

Latency can be tricky... The best solution is to monitor yourself via the keyboard's analog output. You won't hear the same VST instrument as the computer, but you will hear the same notes & timing.

There are input & output buffers (digital "storage tanks" that act as delays). These allow the sound to come-in and go-out at a smooth constant rate, while the computer multi-tasks. (PCs are always multi-tasking even when you are running one application. With MIDI, you are bypassing the audio input buffer but you still have an audio output buffer. A professional interface that comes with ASIO drivers can usually have lower latency (it will work with smaller buffers) than an interface with only Windows Drivers. There is also a "universal" driver called ASIO4ALL that works along with Windows drivers and most hardware. ASIO4ALL still uses some Windows drivers, but it may make a latency improvement.

Your MIDI software and library may also add some latency. I don't know if there is any way to minimize that delay. I'm not a MIDI expert and I don't know if any professional musicians use VST instruments through a computer for live performance.

This post has been edited by DVDdoug: Jun 24 2013, 20:12
Go to the top of the page
+Quote Post
phofman
post Jun 25 2013, 15:15
Post #4





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



I played with stock winxp, midi usb keyboard and envy24 soundcard using regular windows drivers (no asio). I could not get the latency down to reasonable level allowing keyboard playback.

I rebooted to ubuntu linux, used lowlatency kernel, fluidsynth with direct raw alsa output with minimum buffer setting and could hardly feel any latency at all.
Go to the top of the page
+Quote Post
punkrockdude
post Jun 25 2013, 15:48
Post #5





Group: Members
Posts: 244
Joined: 21-February 05
Member No.: 20022



QUOTE (phofman @ Jun 25 2013, 15:15) *
I rebooted to ubuntu linux, used lowlatency kernel, fluidsynth with direct raw alsa output with minimum buffer setting and could hardly feel any latency at all.
Linux is amazing for consumer (and a few prosumer) sound cards and music making.
Go to the top of the page
+Quote Post
[JAZ]
post Jun 25 2013, 19:11
Post #6





Group: Members
Posts: 1783
Joined: 24-June 02
From: Catalunya(Spain)
Member No.: 2383



QUOTE (phofman @ Jun 25 2013, 16:15) *
I played with stock winxp, midi usb keyboard and envy24 soundcard using regular windows drivers (no asio). I could not get the latency down to reasonable level allowing keyboard playback.

I rebooted to ubuntu linux, used lowlatency kernel, fluidsynth with direct raw alsa output with minimum buffer setting and could hardly feel any latency at all.


So you compared like a children's car to a real car? Latencies of 5 milliseconds and lower have been possible in windows for more than 10 years. If you avoid to use the tools that offer that, then, you should do it too in linux and use ALSA with DMIX, or pulseaudio. Don't you think so?

This post has been edited by [JAZ]: Jun 25 2013, 19:11
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Jun 26 2013, 13:34
Post #7





Group: Members
Posts: 3797
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (phofman @ Jun 25 2013, 10:15) *
I played with stock winxp, midi usb keyboard and envy24 soundcard using regular windows drivers (no asio). I could not get the latency down to reasonable level allowing keyboard playback.

I rebooted to ubuntu linux, used lowlatency kernel, fluidsynth with direct raw alsa output with minimum buffer setting and could hardly feel any latency at all.


You apparently need to learn how to spell Asio so that you can compare apples to apples. I'm under the impression that there are some great Asio drivers for Envy24 and Windows.
Go to the top of the page
+Quote Post
phofman
post Jun 26 2013, 21:12
Post #8





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



QUOTE ([JAZ] @ Jun 25 2013, 20:11) *

So you compared like a children's car to a real car? Latencies of 5 milliseconds and lower have been possible in windows for more than 10 years. If you avoid to use the tools that offer that, then, you should do it too in linux and use ALSA with DMIX, or pulseaudio. Don't you think so?


No, I did not have to search for any low-latency mixer-avoiding drivers supporting my card in linux, all the required tools were just there. In fact I used stock kernel later on with the same result.

I can imagine it would have worked OK in windows, just did not want to mess with the system. No windows-linux fight here, just saying what I did.
Go to the top of the page
+Quote Post
phofman
post Jun 26 2013, 21:25
Post #9





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



QUOTE (Arnold B. Krueger @ Jun 26 2013, 14:34) *
You apparently need to learn how to spell Asio so that you can compare apples to apples. I'm under the impression that there are some great Asio drivers for Envy24 and Windows.


Are there any generic Envy24 asio drivers? E.g. working with my http://product.pconline.com.cn/pdlib/34291...ure2462012.html ? Many envy24 cards require specific codec configs, i.e. drivers, a generic driver will not work properly. I know what I am talking about http://git.alsa-project.org/?p=alsa-kmirro...;s=pavel+hofman .
Go to the top of the page
+Quote Post
bennetng
post Jun 27 2013, 10:28
Post #10





Group: Members
Posts: 224
Joined: 22-December 05
Member No.: 26587



Windows ME/2000/XP can use KS, Vista and later can use WASAPI to achieve low latency without using ASIO. DAWs like Sonar and Reaper support them. Generic ASIO drivers like ASIO4all also work well, including onboard audio interfaces and soundcards like my sound blaster cards and onboard VIA HD audio devices.

If Envy24 cards don't work well then it is just Envy24's problem.
Go to the top of the page
+Quote Post
phofman
post Jun 27 2013, 10:51
Post #11





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



QUOTE (bennetng @ Jun 27 2013, 11:28) *
Generic ASIO drivers like ASIO4all also work well


Does ASIO4All offer low latency too? I do not know exactly where they are hooked into the chain and if they can ask the driver for buffer of specific length or they just offer variable length to the calling application but use some fixed length one provided by the windows driver.
Go to the top of the page
+Quote Post
bennetng
post Jun 27 2013, 11:41
Post #12





Group: Members
Posts: 224
Joined: 22-December 05
Member No.: 26587



QUOTE (phofman @ Jun 27 2013, 17:51) *
QUOTE (bennetng @ Jun 27 2013, 11:28) *
Generic ASIO drivers like ASIO4all also work well


Does ASIO4All offer low latency too? I do not know exactly where they are hooked into the chain and if they can ask the driver for buffer of specific length or they just offer variable length to the calling application but use some fixed length one provided by the windows driver.


Instead of thinking about those theories it is better to actually install and try it, if you only use Linux and don't have Windows just do a search and find some screenshots and discussions.

http://img.afreecodec.com/top/screenshot/a...1250587869.jpeg
http://forum.cockos.com/showthread.php?t=49327
Go to the top of the page
+Quote Post
phofman
post Jun 27 2013, 12:30
Post #13





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



QUOTE (bennetng @ Jun 27 2013, 12:41) *
Instead of thinking about those theories it is better to actually install and try it, if you only use Linux and don't have Windows just do a search and find some screenshots and discussions.

http://img.afreecodec.com/top/screenshot/a...1250587869.jpeg
http://forum.cockos.com/showthread.php?t=49327


Looking at the asio4all GUI I recall I tried it (windows XP + cakewalk + usb midi keyboard) though I did not measure the actual latency. I tried to reduce the latency in the asio4all window but it was too large for reasonable playback.


The GUI interface does not explain what each option does. I am not talking about theory but about how the software/hardware actually works. Which is what not many people care about, but I do. That is why I am using open source, I do not like blind guesses smile.gif

Those linked measurements do give real numbers, good. They show there is a rather large fixed latency the asio4all cannot reduce, it most likely uses additional buffer layer on top of a fixed buffer provided by the original windows driver.

My try was not successful, the figures in the link about asio4all latency measurements list values starting at around 10ms. Is anyone using asio4all successfully (i.e. the overall latency at levels acceptable for viable midi keyboard playback)? There are varying reports

http://forums.dv247.com/recording-mixing/9...y.html#post5904

http://forums.dv247.com/recording-mixing/9...y.html#post6226

I am just saying it was quite easy to setup unnoticeable midi keyboard latency in stock linux for my wife to use the keyboard as a regular piano to accompany singers she is teaching. While my relatively limited effort (stock asio4all, soundcard with no native asio drivers, no fine tuning midi USB input (?) ) was unsuccessful in win XP. The hardware was ancient AMD duron 700MHz.

My conclusion for the OP (which the whole discussion relates to) is yes, midi keyboard playback is relatively easily achievable even on weak hardware and any soundcard, at least in linux.
Go to the top of the page
+Quote Post
bennetng
post Jun 27 2013, 13:25
Post #14





Group: Members
Posts: 224
Joined: 22-December 05
Member No.: 26587



The values showed in the Reaper forum are round trip (loopback) latencies, loopback is not needed in playing software synths so actual value should be halfed (10ms -> 5ms for example).

There is no noticeable latencies in my PCs even when playing 10+GB EWQL libraries using keyboard in XP and Win7.

Also, USB devices usually have higher latenceis than firewire and PCI(e) and they are hardware issues rather than OS issues.

This post has been edited by bennetng: Jun 27 2013, 13:35
Go to the top of the page
+Quote Post
phofman
post Jun 27 2013, 15:57
Post #15





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



QUOTE (bennetng @ Jun 27 2013, 14:25) *
The values showed in the Reaper forum are round trip (loopback) latencies, loopback is not needed in playing software synths so actual value should be halfed (10ms -> 5ms for example).


512 samples buffer /44.1 kHz = 11 ms lowest possible latency. The first table lists 17ms. Does not sound like roundtrip time to me.

QUOTE
There is no noticeable latencies in my PCs even when playing 10+GB EWQL libraries using keyboard in XP and Win7.


Asio4all (i.e. the shim over non-asio driver) or native asio?

QUOTE
Also, USB devices usually have higher latenceis than firewire and PCI(e) and they are hardware issues rather than OS issues.


Yes, a very nice test of USB vs. PCI latency: http://apps.linuxaudio.org/wiki/jack_latency_tests
Go to the top of the page
+Quote Post
phofman
post Jun 27 2013, 16:19
Post #16





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



QUOTE (phofman @ Jun 27 2013, 16:57) *
512 samples buffer /44.1 kHz = 11 ms lowest possible latency. The first table lists 17ms. Does not sound like roundtrip time to me.


Actually if it is the ASIO buffer which by ASIO definition always contains 2 periods, then minimum possible latency is slightly above one period, i.e. 6ms. It could be roundtrip then.
Go to the top of the page
+Quote Post
bennetng
post Jun 27 2013, 18:06
Post #17





Group: Members
Posts: 224
Joined: 22-December 05
Member No.: 26587



The same card using native driver and asio4all
http://youtu.be/s_utlvX4xCs

PS: 64 samples work fine in audio playback such as foobar2000, but playing large virtual instruments I need around 96. Native driver has benefits like multiclient capability and more flexible signal routing functionalities though.
Go to the top of the page
+Quote Post
phofman
post Jun 27 2013, 18:26
Post #18





Group: Members
Posts: 304
Joined: 14-February 12
Member No.: 97162



Thanks for the test. If I read the figures correctly the latency of ASIO4All2/windows driver combo for your card is comparable to native ASIO drivers.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 22nd September 2014 - 06:05