Why does cheap integrated sound have better latency than expensive audio interface..?

Configure and optimize you computer for Audio.
RELATED
PRODUCTS

Post

Can somebody explain me that how is it possible that you get much better latency by using integrated soundcard of a laptop than with expensive audio interface?

I just tried, with ASIO4ALL audio interface I can use 64 sample buffer without glitches even with high CPU load (at least in a short test). But with audio interface I get glitches even with 128 sample buffer and I have done all possible optimizations.

At this stage people start to talk about the quality of drivers. in USB you have to buy 550eur RME Babyface to get decent quality drivers and low latency. Or you can use integrated sound that costs probably less than 5-10eur and has much less man-work-hours put into developing its drivers.



According to my understanding the problem is that even with optimized configuration there is always something else that can get higher priority than your audio interface, but with built-in soundcard it always gets a high priority.

I just don't understand why this is so difficult to fix. Isn't there any way to say your OS that ANYTHING that device in USB port number XYZ needs is always in higher priority than any other process?




I mean the situation is just ridiculous. I don't know any other computer device that would have nearly as many problems as audio interfaces. You go and buy a printer, it usually works without problems. The same with a scanner or a web camera or USB stick, USB mouse, etc..etc.. But audio interface ALWAYS some problems. I read complaints even that RME has compability problems with something and it is considered in many places the most stable audio interface brand.

Post

Im sure youve posted on this before, if not, its a familiar theme. Off the shelf Laptops are notorious for low latency audio trouble. Realtime audio, like realtime video is a specialist area. Its never wise to think that you can just buy any Laptop without making sure its meant to run specialist software. You cant buy pro or semi pro soundcards or audio software in high street shops for the same reason. Its not what they are meant for. You cant use a Ford to win Formula one, you buy the right tool for the job. Off the shelf laptops are not sold for realtime specific tasks. Its a gamble if it will work ok or not. Sometimes you can do simple things like disable/change antivirus or wifi, turn off power saving and things will be fine. In other cases, nothing will help. There are Laptops designed for realtime audio/video, if its important then you buy one of those.

Post

What audio interface specifically are you using? Tried that one with Asio4All instead of it's own Asio driver as well?
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post

golemus wrote: I just tried, with ASIO4ALL audio interface I can use 64 sample buffer without glitches even with high CPU load (at least in a short test). But with audio interface I get glitches even with 128 sample buffer and I have done all possible optimizations.
Sample buffer and latency are not absolutes.

You can set a buffer to 128 on one device and see a great 9µs RTL score and then on another you could set it to a 64 buffer and end up with a completely unusable 16µs.

Just because you set A4A to 64 buffer samples it isn't a cert that it would offer up more performance than another dedicated device running at 256 with better drivers.
golemus wrote: At this stage people start to talk about the quality of drivers. in USB you have to buy 550eur RME Babyface to get decent quality drivers and low latency. Or you can use integrated sound that costs probably less than 5-10eur and has much less man-work-hours put into developing its drivers.
The Scarlett for under 100eur would most likely walk all over the onboard using A4A in regards to noise levels, signal quality and driver performance.
golemus wrote: According to my understanding the problem is that even with optimized configuration there is always something else that can get higher priority than your audio interface, but with built-in soundcard it always gets a high priority.

I just don't understand why this is so difficult to fix. Isn't there any way to say your OS that ANYTHING that device in USB port number XYZ needs is always in higher priority than any other process?
If every thing has it's own chip then yeah. In systems where you get chips that take care of 4 or 5 different tasks at the same time then no, your at the mercy of the designer.
golemus wrote: I mean the situation is just ridiculous. I don't know any other computer device that would have nearly as many problems as audio interfaces. You go and buy a printer, it usually works without problems.
For the first week ;)
(I hate printers)
golemus wrote: The same with a scanner or a web camera or USB stick, USB mouse, etc..etc.. But audio interface ALWAYS some problems. I read complaints even that RME has compability problems with something and it is considered in many places the most stable audio interface brand.
Because none of those require high amounts of data being shoved through in real time. If you drop a data bit when you move the mouse then the other few hundred you send ever second will correct it without you realising. If you drop a bit from your audio signal it then goes pop as it needs to be processed in real time. The buffer your adjusting is how much margin in given before it all goes to pot in audio terms, but if your wifi or something decides to grab the CPU for a few cycles that buffer is eaten up really quickly and things go to hell. The problem also exists to some extent with other forms of realtime processing it's just most of those are far more neiche than audio work and even then those are the kind of markets where people are spending real money on powerful boxes rather than trying to re-purpose a word processor bought from best buy.

Using the example of the RME I'd argue that their kit meets offical ISO spec, so the compatability issue would be that the device its getting plugged into not being compatible with the RME and not the other way around.

Post

Anyone who is serious about audio and running mid-to-high-end kit like an RME interface is going to build / invest in a system that is designed to work well with this interface. To use a car analogy ...

There is little point buying a F1 engine and putting it in a Mini. It might be fun for a while, but you're not going to win any races.

Peace,
Andy.
... space is the place ...

Post

golemus wrote:I mean the situation is just ridiculous. I don't know any other computer device that would have nearly as many problems as audio interfaces.
How many other computer devices do you know of that have the same constraints of needing to output a bit-perfect, sample-accurate signal with an absolute minimum bandwidth of around 175 kilobytes per second (for 441.Khz 16 bit stereo)?
my other modular synth is a bugbrand

Post

whyterabbyt wrote:
golemus wrote:I mean the situation is just ridiculous. I don't know any other computer device that would have nearly as many problems as audio interfaces.
How many other computer devices do you know of that have the same constraints of needing to output a bit-perfect, sample-accurate signal with an absolute minimum bandwidth of around 175 kilobytes per second (for 441.Khz 16 bit stereo)?
Of common computer devices maybe display adapters...? Except that the bandwidth requirement is even much higher. And I am not talking only about dropouts/gliches but also of more serious problems like BSOD, sequencer crash, audio going to total distortion mode, etc.etc...

Post

golemus wrote:
whyterabbyt wrote:
golemus wrote:I mean the situation is just ridiculous. I don't know any other computer device that would have nearly as many problems as audio interfaces.
How many other computer devices do you know of that have the same constraints of needing to output a bit-perfect, sample-accurate signal with an absolute minimum bandwidth of around 175 kilobytes per second (for 441.Khz 16 bit stereo)?
Of common computer devices maybe display adapters...? Except that the bandwidth requirement is even much higher.
no, its not sample-accurate; there's a higher bandwidth, but the data doesnt need to be delivered at a fixed, constant, accurate rate.
and your eye doesnt pick up discontinuities the way your ear does; if it did, film wouldnt work.
my other modular synth is a bugbrand

Post

well, we'll be in such [bad] situation until Microsoft decides to incorporate something similar to CoreAudio into Windows platform. And because audio market is a niche, it's highly unlikely that its going to happen anytime soon. apparently both Apple and Microsoft seem to be forgetting about their pro users recently

Post

Tehnik wrote:well, we'll be in such [bad] situation until Microsoft decides to incorporate something similar to CoreAudio into Windows platform. And because audio market is a niche, it's highly unlikely that its going to happen anytime soon. apparently both Apple and Microsoft seem to be forgetting about their pro users recently
To be fair, Microsoft's view of what a 'pro user' is is somewhat different from yours, I think. Businesses are their core pro users, and always have been. Everything on the media side is almost a side effect of the PC as a games platform. Apple, however, do seem to be drawing further and further away from their stalwart 'content creation' loyalists. Im genuinely surprised at how much dissent/annoyance Im hearing from folk 'round here who not so long ago were Apple diehards.
my other modular synth is a bugbrand

Post

Mainstream = more users = higher chances for better drivers.
A soundcard is more software than hardware to me, drivers are vital.
DOLPH WILL PWNZ0R J00r LAWZ!!!!

Post

tony tony chopper wrote:Mainstream = more users = higher chances for better drivers.
A soundcard is more software than hardware to me, drivers are vital.
Well look at Native Instruments, they have sold probably much more audio interfaces than any other vendor in the last 1-2 years (most of them in the form of DJ controller though), and they are still unable to make drivers that are trouble free.

The same with M-Audio. Fasttrack Pro was in TOP-10 sold audio interfaces in Thomann for like 1-2 years or even longer and the drivers (at least for Windows 7) are horrible piece of shit causing a LOT of trouble.

Post

Well look at Native Instruments, they have sold probably much more audio interfaces than any other vendor in the last 1-2 years (most of them in the form of DJ controller though), and they are still unable to make drivers that are trouble free.
True that their drivers aren't the best (while their controller editor is really a model out there, their Tracktor F1 still won't work here), but NI isn't exactly mainstream, it targets a niche market whereas Creative is mainstream.

I don't want anything but an X-Fi in my PC, and it's not for the hardware at all, it's because it has very solid, unbreakable drivers (important when programming), and fully multiclient on top of it.
DOLPH WILL PWNZ0R J00r LAWZ!!!!

Post

I have an x-fi card in my opinion it doesn't hold a candle to my steinberg MR816 fire wire interface in terms of low latency and stability.

Best thing to do is transform your OS around making music only. Windows or Mac to many background processes and routines can disrupt low latency audio processing.

Also recording with these two interfaces is night and day. Really you need an interface that isn't made in china that's all there is to it.
The sleeper must awaken.

Post

diggler wrote:Really you need an interface that isn't made in china that's all there is to it.
If you take an RME card and strip off all components made in China, what's left? An empty cardboard box printed in the EU maybe? :lol:
We are the KVR collective. Resistance is futile. You will be assimilated. Image
My MusicCalc is served over https!!

Post Reply

Return to “Computer Setup and System Configuration”