Latency

Discussions about the use and operation of SAC (Software Audio Console)

Latency

Postby jlepore » Thu Dec 04, 2014 10:57 pm

While I was going to post a thread on some latency measurements I took today, the opportunity to point out some more bullshit being currently spewed by "god" on the other board has presented itself:

Re: Lowering latency

--------------------------------------------------------------------------------

SAC is certainly written to use multiple processors... this has been mis-quoted many times... but as has been stated in many of my messages, there are reasons why multiple processors can actually cause a loss in performance for a realtime engine such as SAC... in reality, the maximum performance can currently be achieved with 2 cores... but using multiple cores can also cause more slipped buffers, depending on many factors.

So... a single highest speed core would give best overall results... with little or no slipped buffers even when loaded to 99%... but in today's world, multiple cores became the norm in the attempt to improve performance and advertise faster and faster computers, since they have pretty much maxed out the speed they can achieve on a single core.

The issue, once again, has to do with a lack of thread priority control once multiple cores are involved as well as a common memory pool across all cores... this is what allows lower priority threads from one core to stomp all over the main realtime priority thread of the engine core, which causes the main thread to not have enough time to finish assembling all of its buffers... overall causing a loss of performance and many slipped buffers.

Bob L


I can challange the entire last paragraph with one word: AMP. It seems to handle all of these problems quite well, and faster than SAC (by far).

From measurements take today:

PC – RME Card – running at 48K in all tests.

Using TotalMix loopback right on the card: 0.79 ms

Using AMP:
(48K / 64 Samples) 0 buff = 4.12 ms 1 buff = 5.46 ms 2 buff = 6.79 ms
(48K / 32 Samples) 0 buff = 2.79 ms 1 buff = 3.46 ms 2 buff = 4.12 ms

Using SAC:
(48K / 64 Samples) 1 buff = 5.46 ms 2 buff = 6.79 ms
(48K / 32 Samples) 1 buff = 3.46 ms 2 buff = 4.12 ms

All measurements were taken using SMAART 7.0 and a SMAART I/O and Pink generated by same.


This system is set to run at 64x2 under SAC (the most stable with only an occasional slipped buffer).
This system runs perfectly at 64x0 under AMP (no slipped buffers)
The latency difference is 6.79 down to 4.12 - a difference of 2.67 ms - It is a significanly noticible difference, even on wedges.
Gigabyte H55-USB3 i5-650/4G/XP/SSD Profire 2626x3/ADAx2/MLA7x1 Motormix x2 AMP/SAC/SAWLite
User avatar
jlepore
 
Posts: 648
Joined: Tue Jun 12, 2012 10:35 am
Location: Minneapolis, MN

Re: Latency

Postby BrentEvans » Fri Dec 05, 2014 12:29 am

Slide the cab closer by a foot. You've just removed 1.1ms of latency. Or slide it back a foot, you just added that much.

It's kind of unfair to compare 2 samples on SAC to 0 Samples on AMP. SAC runs perfectly well on 1x latency settings. In all of the various iterations of my SAC I've used, increasing to more buffers has NEVER fixed a stability problem. Increasing sample size has, but not simply increasing the latency. The fact is that if you're dropping buffers often enough to need that second sample, you're probably dropping more than one buffer at a time, and that one extra isn't going to help. That said, compared fairly, the difference would be either 0.67ms (32) or 1.3ms (64), and I can't see where that is audible. Most performers move that much in the sound field during their performance. 2.67, maybe, if you were listening critically and had some observer bias...

If we're talking about in-ears, certainly having lower latency can make a difference in the phase cancellation a performer hears, but that's also true of increasing latency, adding reverb, and changing EQ (using normal EQ algorithms that aren't linear phase).

I'm not arguing that AMP is superior in its capabilities, or that SAC is based on seemingly obsolete algorithms... but when we get to this level we're talking metaphorical dancing angels and the heads of metaphorical pins.
User avatar
BrentEvans
 
Posts: 269
Joined: Thu Jun 21, 2012 5:05 pm

Re: Latency

Postby jlepore » Fri Dec 05, 2014 2:14 am

No, actually the difference in the real world is startling. Performers notice it immediately, I notice it immediately in the monitor EQ's, and monitoring through a headset and hearing your talkback in your head is not something that makes you go mad.

This is the EXACT same machine that was running SAC and needed to run at 2x64. I have now run 4 successful monitor gigs (under real show pressure on national act days) and it was FLAWLESS at 0x64. The fact that SAC can't do 0x64 is based on the assumptions and excuses of Bob, not reality.
Gigabyte H55-USB3 i5-650/4G/XP/SSD Profire 2626x3/ADAx2/MLA7x1 Motormix x2 AMP/SAC/SAWLite
User avatar
jlepore
 
Posts: 648
Joined: Tue Jun 12, 2012 10:35 am
Location: Minneapolis, MN

Re: Latency

Postby RBIngraham » Fri Dec 05, 2014 12:02 pm

BrentEvans wrote:Slide the cab closer by a foot. You've just removed 1.1ms of latency. Or slide it back a foot, you just added that much.

It's kind of unfair to compare 2 samples on SAC to 0 Samples on AMP. SAC runs perfectly well on 1x latency settings. In all of the various iterations of my SAC I've used, increasing to more buffers has NEVER fixed a stability problem. Increasing sample size has, but not simply increasing the latency. The fact is that if you're dropping buffers often enough to need that second sample, you're probably dropping more than one buffer at a time, and that one extra isn't going to help. That said, compared fairly, the difference would be either 0.67ms (32) or 1.3ms (64), and I can't see where that is audible. Most performers move that much in the sound field during their performance. 2.67, maybe, if you were listening critically and had some observer bias...

If we're talking about in-ears, certainly having lower latency can make a difference in the phase cancellation a performer hears, but that's also true of increasing latency, adding reverb, and changing EQ (using normal EQ algorithms that aren't linear phase).

I'm not arguing that AMP is superior in its capabilities, or that SAC is based on seemingly obsolete algorithms... but when we get to this level we're talking metaphorical dancing angels and the heads of metaphorical pins.


Brent, with all due respect you always go to the fact that moving a wedge a foot or two equates to the same change in latency and that is true of course. However there are plenty of real world situations where knocking off another ms or two can make a big difference. Obviously I always think about if from a theatrical perspective but that is a good example, your typical center cluster of speakers is always too latent, even with a near zero latency all analog system, the sound from the actors mouth is arriving before the sound from the speakers and you can not put a time alligner on the actors mouth. :) In any situation like that, the less the better.

Also to me it's worth pointing out Bob L's real bullshit here and that is continuance of the multiple core nonsense. Compare SAC to AMP on any system with two or more cores and you'll see SAC put most if not all of it's load on one of the cores, where as AMP will distribute itself pretty evenly. I'm not savvy enough about programming and computer architecture to say why... but it's pretty clear and at least to me it has been pretty clear for years now, that Bob is full of shit on this subject. Many DAWs have used multi-core fairly efficiently for years now... What he does or does not do that makes his stuff different... no idea... but I see the same results whether it's looking at DAWs, AMP vs SAC, and even in the software I use for theatre playback and such...
Richard B. Ingraham
RBI Computers and Audio
http://www.rbicompaudio.20m.com/
SAC details and goodies at: http://www.rbicompaudio.20m.com/SAC.html
RBIngraham
 
Posts: 1021
Joined: Tue Jun 12, 2012 1:05 pm
Location: Cleveland, OH USA

Re: Latency

Postby Jeffsco » Fri Dec 05, 2014 12:22 pm

The Latency, especially with Edrums, running SAC at 1x64 is just "usable" with one of the Drummers I work with. 2x64 was iffy at best. I'm considering giving him a Drum submix derived from my Raydat's Totalmix so as to eliminate a few more ms from the mix.

Perhaps that would work with In Ears as well? The basic feed is sent by Totalmix and then somehow add the reverb or other instuments / vocals back in with SAC processing?

Don;t know how that would be routed.

Too bad the Raydat cards do not have the Totalmix FX set up. I've yet to upgrade my card to the latest firmwre build but it should make using it much easier.
Jeffsco
 
Posts: 102
Joined: Tue Jun 19, 2012 10:32 am

Re: Latency

Postby shmick » Fri Dec 05, 2014 1:25 pm

TotalMix FX will run on the RayDAT

At least according to http://www.rme-audio.de/download/window ... er_4_e.pdf
Host: ASUS P5Q SE/R, Intel E8400 O/C'd to 3.8ghz, 3 x RME HDSP 9652, XP Pro
Gear: 9 x ADA8K, 4 x Audiorails, 1 x BCF2000
Config: FOH + 12 stereo IEM mixes
Misc: Dual Linkwitz-Riley plugin, Studio Levelizer, Studio Reverb, Frequency Analyzer, SAWStudioLite
User avatar
shmick
 
Posts: 359
Joined: Tue Jun 12, 2012 1:55 pm
Location: Toronto

Re: Latency

Postby Jeffsco » Fri Dec 05, 2014 2:33 pm

I realize that Totalmix FX runs on the Raydat card....it just doesn't have any of the FX. They've reconfigured the mixer GUI and routing....which is good news as the original GUI took a bit to figure out. They can't include the FX as the Raydat Chipset does not have that ability.
Jeffsco
 
Posts: 102
Joined: Tue Jun 19, 2012 10:32 am

Re: Latency

Postby lowdbrent » Sat Dec 06, 2014 1:29 am

You can't slide your IEMs back 4 feet. 4 to 6ms is unacceptable. Modern systems are 2ms or less micro seconds round trip. Some are 200 micro seconds.
lowdbrent
 
Posts: 71
Joined: Sun Aug 05, 2012 5:54 pm

Re: Latency

Postby RBIngraham » Sat Dec 06, 2014 9:29 pm

lowdbrent wrote:You can't slide your IEMs back 4 feet. 4 to 6ms is unacceptable. Modern systems are 2ms or less micro seconds round trip. Some are 200 micro seconds.


OK... I'll bite... what console has a round trip of only 200 micro seconds? I aausme that would be at 96K?
Richard B. Ingraham
RBI Computers and Audio
http://www.rbicompaudio.20m.com/
SAC details and goodies at: http://www.rbicompaudio.20m.com/SAC.html
RBIngraham
 
Posts: 1021
Joined: Tue Jun 12, 2012 1:05 pm
Location: Cleveland, OH USA

Re: Latency

Postby Butch » Sun Dec 07, 2014 1:16 am

And I thought the SSL was quick at about 1ms
So ya what has 200us
using 1 64 in SAC has worked good for me

Butch
Butch
 
Posts: 107
Joined: Wed Jul 11, 2012 11:41 pm

Next

Return to SAC Discussion

Who is online

Users browsing this forum: No registered users and 16 guests

cron