while some can be "tweaked" to be sort of ok (might be ok for lite duty and @ 512 buffer)
 most can not
it requires a custom bios. that has to come from the manufacturer and the only reason mine work well.

some of the tweaks are crazy like turning off the video card drivers and using VGA mode (very ugly)
also disabling the DVD-RW (for the off the shelf ones)
and again USB has a far better change of working..

jcschild


while some can be "tweaked" to be sort of ok (might be ok for lite duty and @ 512 buffer)
 most can not
it requires a custom bios. that has to come from the manufacturer and the only reason mine work well.

some of the tweaks are crazy like turning off the video card drivers and using VGA mode (very ugly)
also disabling the DVD-RW (for the off the shelf ones)
and again USB has a far better change of working..

And don't forget the real tech support!

pinguinotuerto


I still "question" the validity or relevance of this DPC Latency test on new machines running Windows 7 with a Core i7 processor.  Like I said before, my DPC results hover between 100+ and 400+, spiking at 500+ and I have no dropouts whatsoever.

Is this a quad core processor?  Does that help with the DPC latency issues?  One core is held up by the DPC, but the others are still free to process audio?  Generally what I've read is that a fast dual core is better than a slower quad core for audio.

normal and tweak is an oxymoron...

just saying.

Is this a quad core processor? Does that help with the DPC latency issues? One core is held up by the DPC, but the others are still free to process audio? Generally what I've read is that a fast dual core is better than a slower quad core for audio.

 
Windows is not technically a realtime OS.
(Even though things happen so fast... it can appear to be realtime)
Everything is scheduled via Deferred Proceedure Calls (thus the acronym DPC)
DPC latency is the time between when those scheduled tasks are executed.
The lower the DPC latency, the closer you are to the ideal of a realtime OS/machine.
 
Note:  The lower you set your ASIO buffer size... the more critical it is to have low/consistent DPC latency.
If you plan to run heavy loads at a 64-sample ASIO buffer size or smaller, low/consistent DPC latency is important.
 
If a particular driver causes high DPC latency (Poor WiFi drivers are notorious for this)... it can cause a hiccup (gap) in data flow.
If this gap happens... and your audio interface's buffer doesn't get filled in time... you'll hear a glitch (pop/tick) or experience a dropout.  The smaller the ASIO buffer size, the greater the odds of buffer under-runs...
 

Jim Roseberry
Everything is scheduled via Deferred Proceedure Calls (thus the acronym DPC)
DPC latency is the time between when those scheduled tasks are executed.
The lower the DPC latency, the closer you are to the ideal of a realtime OS/machine.

Yep.

"A device driver cannot process data immediately in its interrupt routine. It has to schedule a Deferred Procedure Call (DPC) which basically is a callback routine that will be called by the operating system as soon as possible. ... There is one DPC queue per CPU available in the system. ... If any DPC runs for an excessive amount of time then other DPCs will be delayed by that amount of time. ... Unfortunately, many existing device drivers do not conform to this advice. Such drivers spend an excessive amount of time in their DPC routines, causing an exceptional large latency for any other driver's DPCs. For a device driver that handles data streams in real-time it is crucial that a DPC scheduled from its interrupt routine is executed before the hardware issues the next interrupt. If the DPC is delayed and runs after the next interrupt occurred, typically a hardware buffer overrun occurs and the flow of data is interrupted. A drop-out occurs."

So I'm asking if a processor with more cores would have fewer problems with DPCs, since other cores could theoretically handle the audio DPCs while one is locked up by a rude Wi-Fi DPC routine.  Is the queue shared between cores, and DPCs can be shuffled around to whichever one is free?  Or is there one queue per core, allowing for a core to be hijacked?  What about virtual cores?

Found something:

http://stackoverflow.com/...pc-latency-testing-how

"A DPC is queued into the global DPC queue, and can be run on any processor. So if you really have a long (-running) DPC on one core, the other core is free to process another. So any timing information is really dependent on the count of processors you have and how many things get currently executed concurrently. So on multicore processors these numbers might vary widely."

mike_mccue


Here's something to consider.... when you are hearing crackles at 96 samples you can expect to find all sorts of difficult to hear errors in audio recordings made during that pass.

You can easily search out and locate these sorts of errors with a app like Wavelab. But it might bum you out to do so.

The kind of folks that hear those tiny errors are the same folks who hear the difference between a 96 sample delay and a 144 sample delay.

Sensitivity to delay is a human thing... you have to work with each musician at their threshold.

Locating errors with an app that features error search functions is more scientific than asking humans what they feel... and if one checks for them when working under different circumstances, one may better understand what their system is truly capable of.

best regards,
mike

Sounds logical.   And a good way to measure it.

I will say this,  I am extremely sensitive to little time differences.  I grew up on drum machines and click tracks.  So I notice right away. 
I also trust my ears.  When I mix down,  I listen very carefully to the mix-down before I send it out.  Several times to be absolutely sure there are no artifacts. 

So that brings me back to what im sort of saying..if you can't hear it or feel it, does it matter?

mike_mccue


"So that brings me back to what im sort of saying..if you can't hear it or feel it, does it matter?"

I think the question/answer becomes more complicated when your guests make the claim that they can hear it... and (I'm speaking only for myself) you know your guests have better musical training, performance experience, and perhaps ears than you do.

best regards,
mike

well in my case,  I am the one being trusted.  I've had a lot of releases and done a lot of recordings for bands.  I don't do the mastering,  so that gets sent off.  no mastering engineer or anybody else has ever said anything.

mike_mccue


on yer contemporary laptop?

Yes.  I've sent several from my laptop already.   But I've done it in all types of computers. Computers are just generic parts. I dont hold any particular reverence towards them.  I love what you can do with them,  but its just generic parts.

oh yeah my msi is way under 500us under vista while under 40us in xp !! it's all about os drivers integration !!

so os is still an issue !!

I doubt Jim will respond to your insult so I will remind you that Jim was testing that particular computer because he himself wants to purchase a new laptop that he can use as his own personal portable DAW.

Some of us feel that we are in a similar situation... It just seems like there are no good laptops to buy.

regards,
mike

smoothjeff


FOUND LAPTOP WITH ACCEPTABLE LATENCY!

I have been reading this thread for a few days while deciding on a laptop to buy and just registered on here to make this post and save some people time. I had a Dell Studio 1535 (look it up it's the most uncurable DPC latency disaster ever), I sold that and bought a Studio 1555 because I read on a forum a guy had no DPC spikes, however the idle DPC was still ranging 150-500us so I was having audio problems. I saw an even faster HP than my Dell for $700 on eBay but went to Costco and it's DPC was ranging from 290-500us. I was about to give up then I saw a review for the Dell Studio 1558 and the reviewer showed a DPC latency test which was fairly decent (compared to other off the shelf laptops) http://www.notebookcheck.net/Review-Dell-Studio-1558-Notebook.24496.0.html
If you look at his link for the Dell 1557 review you can see that one had the same crap DPC levels, but the 1558 was acceptable so I looked specifically to try another Dell with the new Intel mainboard and i5 cpu. Heres the link for the laptop I bought (it's on sale until January 30th for $649)
http://www.staples.com/Dell-Insprion-17.3-Laptop-(Blue)/product_851069

I didn't like that it didn't have a graphics adapter with dedicated memory, plus I'm going to have to upgrade the 5,400rpm hard drive for something faster, but this is the first time I've been able to use Guitar Rig at low latency without any problems. The latencies range around 60-140us on average, but this is with wifi on and straight out of the box with all of Dell's bs on it. I'm hoping I can get it consistently lower with a fresh install of Windows and some tweaking if anyone has any good suggestions to save me time?

It's not as low as the computer in Jim's video, and probably won't handle as much as a good desktop but for $649 I'm glad I can get something out of this. And considering I took my flash drive with the DPC latency checker and copied it onto every semi-powerful laptop in Best Buy and Staples, and the one I bought was the ONLY one with consistent latency under 130us, I didn't have much of an option. I took videos of several laptops with the DPC latency checker (with wifi turned off) but here's the one I bought (I used the camera on my new phone so it's really hard to see but it's ranging from 60-100us in the video)
http://www.youtube.com/watch?v=jCqJ0JhuFto


Again, I've already put too much time into finding a laptop vs. actually playing/recording the music it's intended for, so if anyone has any tweaks to try out to get the DPC latency lower, that would be appreciated

Does anyone know which Dell laptop smoothjeff bought? Both links are dead.
A 17" screen would be good and he seemed to find a decent DPC latency