PRO TOOLS HD LATENCY VS. LOGIC 7 PRO

[Tetness]

I am buying PTHD3 this week to use beneath my Logic Pro because I want as little latency as possible, among other benifits. I have been searching to determine exactly how much better PT latency is over Logic Pro with a G5 dual 2.5. Can someone explain me the difference? Thanks

[T-fonk]

Save your money and buy a RME card with DSP for routing! Like the old "Multiface" or something like that! Don't get the HD-systems if you're a logic user only! And for certain not if your issues is latency only!

T-fonk

[studjo]

T-fonk I'm sure you work everyday with PT - right?

Jo

[T-fonk]

Originally Posted by studjo

T-fonk I'm sure you work everyday with PT - right?

Jo

Yes. Everyday at my HD 2 Accel. Why you ask?

T

[Tetness]

Currently I am running Logic Pro 7.1 with a Lavry Blue and a Lynx AES 16. My question now is, will getting PTHD improve my latency and how much. Also, I wouldn't mind having more plug-in processing power. Money's not really the issue here, so would getting PT be slutty, or just plain stupid?

[rolo]

as said before.... there is no such thing as latency with an hd system. go to www.digidesign.com to educate yourself on the hd line. if you understood that there was latency on hd... there may be many other things you are unaware of on HD systems.

[Tetness]

Cool!

[XHipHop]

you mean tracking latency, not plugin latency, right?

[Tetness]

I'm talking about tracking latency. The time it takes from my guitar pick to send a signal through my guitar into my mic pre, into my compressor, into my lavry Blue, into my computer, into Guitar Rig and out of the computer to my lavry and into my speakers. The difference between PTHD and Logic Pro.

Also, I just had a conversation with a friend who told me that Neundo has hardly any latency. Much bettter than Logic Pro. What is the reality of this?

[doug_hti]

Both platforms will have acceptable latency while tracking smaller sessions.

Part of the difference comes when you have a large session with existing plugins and you are tracking something, you will never have to fudge buffer settings or deactivate plugins to have descent latency if monitoring through the AD/DA.

Depending on what the lavry's latency is, your latency will be in the neighborhood of around 200 samples give or take 50 with HD in to out.

[logichead]

Originally Posted by Tetness

I'm talking about tracking latency. The time it takes from my guitar pick to send a signal through my guitar into my mic pre, into my compressor, into my lavry Blue, into my computer, into Guitar Rig and out of the computer to my lavry and into my speakers. The difference between PTHD and Logic Pro.

Also, I just had a conversation with a friend who told me that Neundo has hardly any latency. Much bettter than Logic Pro. What is the reality of this?

There is some latency when tracking through HD. It's not very much, but no question - if you split the input before it goes to the HD rig and mult it into a console you can hear and feel the difference. Put both signals together and you will hear flanging that doesn't move.

Certain plugs add latency if you are monitoring through them.

I use Logic as the front end to an HD2Accel / Dual 2.5 as well as a native rig. My old O2R serves as input monitoring for the native rig and it is noticeably "quicker" than the HD rig.

I also have checked this on the HD rig, monitoring via the O2R and comparing the same signal from both the O2R and the HD. The amount of latency we are talking about is very small - like being 10 feet away from your guitar amp - but it's there.

There are plenty of reasons to get an HD rig - but these days the distance between tdm and native is smaller than ever.

Best....H

[Tetness]

Logichead, thank you. Your reply was very informative. thumbsup

[T-fonk]

But if you wan't to play through NI's guitar rig, I'll guess it's about the same latency on all systems! Unless guitar rig comes as a TDM-plug! I guess that will never happen...

T-fonk

[kosi]

And btw: There's no Delay compensation, when you use it as frontend for PT !
This sucks hard...

[nativeaudio]

Originally Posted by Tetness

I am buying PTHD3 this week to use beneath my Logic Pro because I want as little latency as possible, among other benifits. I have been searching to determine exactly how much better PT latency is over Logic Pro with a G5 dual 2.5. Can someone explain me the difference? Thanks

PT HD tracking latency @ 44.1 kHz: 105 samples (2.3 ms, hardly audible in most cases).
Native OSX systems at 64 samples buffer: normally around 6 ms on a computer than can handle the 64 samples setting without introducing audio artifacts.

You could possibly buy a HD1 (or a little, good mixer) until Apple comes up with a solution (like an advanced I/O), which hopefully won't take too long.

[tomdarude]

I´d skip the HD system, of course if ya rely on the HD192 and PT alone it´s a nice and safe solution, but.....if you only use logic on top of PT and...

if you add highend converters you add latency! an amazing A/D needs some time to do it´s wonders....so if you add lavry´s / prism´s and skip the digi-cons, I wonder what latency you´ll measure ????

I´d say a G5 with a lynx aes16 & lavry blue are a really HOT system, I´d spend the money on a nice ANALOG monitoring solution, that has NO latency vs. 2-3ms (digi)
stay tuned for the new dangerous monitor controller&cue system (will be around 1,5k only!!) you´ll get a highend monitoring solution, a cuesystem with a nice headphone out and talkback. and zero latency monitoring !!without having to worry about the next software/driver-update etc.
that´s overall soundquality improvement as long as your not already in this league (monitoring wise...read: avocet etc.)
and with really nice guitar-/bassplayers we DID notice some changes in their playing with analog direct monitoring !! (using PT HD!!) you won´t hear the 2ms..the player won´t really notice it ...BUT he was definetly playing tighter/better with direct monitors..... it´s not about hearing it´s about feeling !!

and you´ll save $ 7500,- !! relying on PT plugins during tracking?? imagine spending 7,5k on a vocal chain (mercury 66´s....wunders.....V76....trakker...old 1176´s...you name it) or get a cranesong spider !!! 8ch. pre-limit-A/D for your lynx

all that will hold it´s value much better than any DAW or computer stuff over the next year´s.... another important aspect, for me at least...

cheers tom

[C.Lambrechts]

Guys .... get your facts ligned up and use them the right way please.

I see things like Pro Tools monitoring having 2 - 3 ms of latency versus other analog analog monitoring systems having no latency ? Care to explain that please ?

Simple fact : data in the computer needs to be translated to your monitors : the DA process .... That's 2 - 3 ms latency Tom is refering to with TDM systems but make no mistake : ANY DA convertor will give you that kind of latency.

Latency is a complex issue and can is way too often used and even more missused to argument one or the other.

In this particular case : tracking.

In tracking at some point you have to go through a AD stage to get into your computer. AD's all average round 2,5 - 3 ms @ 48k. That is the time the convertor needs to translate the analog signal to a digital signal. No way around that wether you use a native or a TDM system. Now , here is the important difference with a TDM and a native system .... what happens AFTER that initial AD latency.

It's the latency that's added at this point that becomes important ... especially since during tracking you have to go back out and listen to what you're doing ... monitoring.

Monitoring adds you're typical 2,5 - 3 ms of DA convertor latency to the signal. Similar to the AD stage.

Now ... this is where TDM systems start to prove what they're worth. Except for the AD / DA process ZERO is added. Meaning true ZERO latency. With native systems .... well .... there's some good ones out there like the rme fireface which are blazing fast. 'doesn't matter' fast even but definately not ZERO latency. Furthermore it might seem irrelevant when you're tracking a single vocal track but rest assured .... another point where TDM still proves what it's worth today is when doing multitrack sessions. TDM doesn't care if you record a single vocal track of a complete band at once doing 24 tracks or even 50 / 60 / 70 tracks at once. It remains ZERO latency. Try that with you're everage native system , fireface or not.

Still not convinced ? I understand .... because even there native has some pretty good solutions which are acceptable .... now instead of starting a fresh session do multi-track over dubs .... have 50 tracks play back through your interfaces ... set up a monitor system for your drummer and your bass player for example and let them play along to the track while listening to the track and themselves. TDM .... zero latyency. Native .... well .... your only solutions is to set up a direct signal monitoring system through a console or something and pray nothing goes wrong. It might work in your average home kitchen appliance studio but there are still professionals out there that prefer to know what they're doing instead of hoping the signal is ok.

There differences between native and a TDM system are getting smaller and less significant every day. Thing is no matter how powerfull your cpu is at some point it needs more time to deal with the information you send it. And if you think 2,5 ms is relevant .... think again .... we're talking 500 and more ms here in some native situations.

Like I said ... the difference might have grown smaller over the last couple of years but it is still VERY significant and still there. And it's not going to go away in the next 2 years either.

Now go ask yourlef if you need all that ... but that's a completely different subject. As is latency introduced by plugin priocessing once the signal is in the daw.

[C.Lambrechts]

Originally Posted by tomdarude

I´d skip the HD system, of course if ya rely on the HD192 and PT alone it´s a nice and safe solution, but.....if you only use logic on top of PT and...

if you add highend converters you add latency! an amazing A/D needs some time to do it´s wonders....so if you add lavry´s / prism´s and skip the digi-cons, I wonder what latency you´ll measure ????

Tom, not sure I'm with you here ? Are you referring to using a HD system in combination with other convertors or simply bypassing the entire HD system ?

Thanks,

[C.Lambrechts]

Originally Posted by logichead

There is some latency when tracking through HD. It's not very much, but no question - if you split the input before it goes to the HD rig and mult it into a console you can hear and feel the difference. Put both signals together and you will hear flanging that doesn't move.

that would be convertor latency and is not due to the HD system. You would get that with ANY system or with ANY convertor. What you're doing is listening to the direct signal through the console and the 'tracked' signal through the HD system. Think record head and playback head for example in the tape days.

Originally Posted by logichead

Certain plugs add latency if you are monitoring through them.

Correct ... again no matter what system you use.

Originally Posted by logichead

I use Logic as the front end to an HD2Accel / Dual 2.5 as well as a native rig. My old O2R serves as input monitoring for the native rig and it is noticeably "quicker" than the HD rig.

I also have checked this on the HD rig, monitoring via the O2R and comparing the same signal from both the O2R and the HD. The amount of latency we are talking about is very small - like being 10 feet away from your guitar amp - but it's there.

How do you monitor / compare ... because it would seem to me that what your saying is absolutely normal. Remember that the O2R is a digital console and that if you feed it a signal (digital or analog) it has to convert (AD or DA or both) before you can actually listen to it. meaning you have to take AD/DA conversion latency into account. Therefore it make absolute sence that there's a difference between listening direct to the HD system or firts going through the O2R


And yet again .... this would occur with ANY system ... not only with HD

[DrDeltaM]

Originally Posted by C.Lambrechts

Still not convinced ? I understand .... because even there native has some pretty good solutions which are acceptable .... now instead of starting a fresh session do multi-track over dubs .... have 50 tracks play back through your interfaces ... set up a monitor system for your drummer and your bass player for example and let them play along to the track while listening to the track and themselves. TDM .... zero latyency. Native .... well .... your only solutions is to set up a direct signal monitoring system through a console or something and pray nothing goes wrong.

About all soundcards for native use have an onboard DSP monitor mixer, which gives you the same zero latency monitoring as ProTools (also a DSP based 'monitoring' system). Of course, this monitoring is without plugins, unless you use a card like the Metric Halo or Creamware stuff.

So advantage of PT is the fact that you can monitor with plugins (selected zero latency ones anyway) and that you deal with only one mixer, while with native you'd have the DAW mixer and the soundcard monitor mixer. But latency itself ain't really the issue. If this difference is worth 7k to you or not is something else

[sysexguy]

whoa, seems like a lot of headaches just to use NI Guitar Rig???????

Andy

[tomdarude]

oh, you want to track through guitar rig ?? ooooops,I´m a really modern guy, BUT I´d say "forget it !!"
I did it, was amazed first, but truly didn´t play that good anymore.I´m serious my playing lacked inspiration, timing and subtile details VS. a miked amp.

what I´m doing right now is a bogner metropolis with a SM57 (no royer121 yet..argh I want one) and a DI. signal .....afterwards the DI. signal gets GuitarRigged/reamped or just some nice pultec treatment and I decide/mix the two...

works out

[Kingtone]

Originally Posted by Tetness

Also, I just had a conversation with a friend who told me that Neundo has hardly any latency. Much bettter than Logic Pro. What is the reality of this?

This is funny... what is it about Nuendo users?... they constantly have to re-assure themselves that they bought the right program!!! every Nuendo owner i know seems to be on a personal mission to diss on Logic or Protools! its Hillarious!!
I figure... if you are that confident that your program is the best... you wouldnt go out of your way to diss another software solution!

I sometimes track on my Logic fronting my HD2 Accel,... but if you are going to be recording using Guitar rig or any software amp plugin... forget it... you will have latency... try using both... mic up an amp and DI your guitar so you can muck around with the plugins later.. but you will still have the response of the real amp.

[tomdarude]

Originally Posted by C.Lambrechts

Tom, not sure I'm with you here ? Are you referring to using a HD system in combination with other convertors or simply bypassing the entire HD system ?

Thanks,

I meant the first, NOT using the digi-converters, but some highend into the digital HD ins....

I heard that from various people building converters....the more highend->the more latency ...makes me wonder what latency a lavry-gold mkIII A/D has ??? or prisms?? anyone??

[tomdarude]

Originally Posted by C.Lambrechts

Guys .... get your facts ligned up and use them the right way please.

I see things like Pro Tools monitoring having 2 - 3 ms of latency versus other analog analog monitoring systems having no latency ? Care to explain that please ?

Think I didn´t make myself clear, sorry....I´m busy with my final architecture-diploma, that has to be finished in 4 weeks and shouldn´t hang around in forums anyway!!! that´s why my brain is pretty fried....rrr......just wanted to take coffeebreak and lurk a bit.....but as ya see...


on topic:
I meant when dealing with a singer or single instrument player....and I meant ANALOG monitoring....get the playback with whatever latency and blend in the analog signal from your recording chain BEFORE the A/D....and then nudge the track by the actual measured latency later... regardless if PT/Native

cheers tom

[sdevino]

Just to put some latency numbers in perspective.

Lets say you booked Abby Roads studio and were using their large tracking room with the whole band setup at the same time.

If you are 20' (about 6 meters) you would hear the drums about 17 ms after the drummer does.

If you are 10' from the bass player you will hear the bass about 9 ms after the bass is played.

If you are a guitar player and use a 20' chord you hear your guitar amp about 17ms after you play each note.

Ok now lets say you monitor through HD then add a really really high latency plugin with 1000 sample latency. At 44.1kHz sample rate the 1000 samples equals 22.7mS, which is about the same as standing 26 feet (8 meters) from your guitar amp. Typical plugins have between 4 and 50 samples. 4 samples is like putting your ear 1.2 inches (3 cm) from your guitar speaker. 50 samples is a little over 1 foot (24 cm).

If you use Chris's numbers for A/D D/A latency they place you an additional 1.2 inches from the speaker.

If you are generally OK playing your guitar within 25 feet of your amp and are not bothered by the latency of time and space, then using an HD rig with all but very few plugins will be fine.

If on the other hand you find you need to stick your head inside the speaker cab or inside the snare drum in order to maintain a groove, then you should worry about latency of TDM rigs.

Steve

[C.Lambrechts]

Originally Posted by tomdarude

I heard that from various people building converters....the more highend->the more latency ...makes me wonder what latency a lavry-gold mkIII A/D has ??? or prisms?? anyone??

yes ... and you're right as far as I know. But it is also relative. The average high quality convertor varies between say 110 and 130 samples at 48k I believe. And yes ... the higher quality ones you mention tend to be on the higher end of those numbers.

[tomdarude]

@ sdevino this is true, BUT.....


that´s where latency-compensation comes into play.....not the one in Logic 7.1 we´ve been waiting for so long ..... , but you´re very own one !!

if you´re standing about 3 meters away from your guitar amp, you´ll have around 9ms latency, right? but there´s some latency compensation going on in your head....
you´re eyes see it´s 3m away and your brain expects the sound to be 9ms late !!
it´s that easy!

(makes me think of a workaround: try GuitarRig running in logic, put your fender amp
5m away and imaging your playing the amp !!!hahaha never tried it actually )

for all people starting to complain that this is peanuts,bullsh** , too little to hear etc.:

close your eyes, some insect is flying around your head, try to catch it blind....got it...okay try again ....got it ,well then.... you did it the other way round...
your ears measured the stereoimage of the insect´s sound and you´re able to HEAR as exact as 10cm, WHERE that beast is flying round your head !!! that equals 0,3 ms !!!


when playing some easy strumming stuff a professional player wouldn´t be bothered with latency, I know....but when it comes down to that inspired moment he´ll be playing better sitting in front of his amp!

love that thread...keep it going.....arrrrghh have to work...nevermind

[nativeaudio]

That's some important comments, tomdarude.

This is from a latency-discussion at osxaudio.com:

The human brain seem to have some built-in, intelligent latency compensation. Of course it cannot compensate for latency, but it pseudo-compensates for the perception of it.

For example, if you speak with a person who is 20 feet away, the latency is 20 ms, you don't feel that there is a latency. The brain links the perception of distance (from the eyes) with the latency, and the viewer/listener gets a feeling that nothing is wrong.

On the other hand, if he would hear his own voice, or the voice of someone 2 feet away, 20 ms late, it would be definitely be registered as 'wrong', just like high latencies on recording systems.

If works the other way too. If someone has been adding sound effects to a movie, and I see a picture of a sudden sound at a long distance, and hear the sound with no latency, it feels wrong.

Plus, there is of course the problem of a singer or drummer or saxophone player who in cases of high latencies will hear both the delayed signal from his headphones in addition to the non-delayed signal from his voice or instrument, which will cause timing/pitch problems.

[Orren]

Originally Posted by Tetness

Also, I just had a conversation with a friend who told me that Neundo has hardly any latency. Much bettter than Logic Pro. What is the reality of this?

Latency is determined by the number of samples between the time a performer hits a note and the time it takes the AD converters + audio buffer (if native, not PTHD) into application + OS/application (if native, not PTHD) + system buffer (native, not PTHD) + DA converters to play that note through the speakers.

The only place that the application comes into play is the "OS/application" stage. In general, any high quality audio application, such as Nuendo or Logic, can do all their required grabbing of audio and writing to disk and spitting it out to the buffer for playback within a single audio buffer, so it will not add any additional samples.

In other words, neither Logic Pro 7 nor Nuendo 3 nor Cubase SX 3 etc. have any different latency from each other. Latency will be determined by all the other parts of the chain.

And as you can see, PTHD has quite a smaller chain, resulting in the nearly imperceptible tracking latency others have noted previously.

Orren