Direct monitoring vs. software monitoring

[nativeaudio]

I've assumed that Logic's default setting would be the best solution in recording situations, and that default is that the signal that is being recorded is being monitored through the software. Of course, direct monitoring means less latency, but I thought that enabling software monitoring would enable something very useful that I would miss if I disabled that function. But is that really true? Why would I want to use software monitoring if I don't need to use inserts on a recording track?

With a 128 buffer setting (x2) and converters, software monitoring means audible latency, often in the 8-10 milliseconds range - even without any plug-ins inserted - and who wants that? :)

[Kiwiburger]

Software monitoring sucks. If your interface doesn't offer zero latency monitoring, use a mixer.

[Beardhead]

For me it has always been out of question to use software monitoring. If you have to use effects on the cans, use outboard gear. It's a no go to influence the musical source in a bad way!
I got the old Roland 480 line mixer thing for that.
Claus

[peeder]

Sometimes latency can be your friend...in fact, sometimes it's useful to have more of it!

Why? For those without control room isolation (or recording themselves), if you monitor the signal with enough latency (e.g. a few seconds) you can completely seperate the sound you're evaluating with the direct signal also getting into your ears. You can also separate it from any visual cues that the end-listener won't be able to appreciate. This is useful for getting your tone printed right.

Anyway 128 samples of latency isn't the end of the world for many situations, but it is too long for recording vocals, where the direct signal in the singer's skull will have nasty phase interactions with that much latency. Low-latency monitoring in PTLE works fast enough on my machine (Mac Pro, 003) to cut vocals with.

Remember that you give up the ability to monitor with plugins active in a low-latency monitoring situation. You can use effects in the analog domain either going in or coming out to solve that. If you go purely in the analog domain for your cue mix, you won't be evaluating the effects of conversion (especially clipping!) if that's what you, the engineer, also are listening with.

[ricfoxx]

Originally Posted by peeder

Sometimes latency can be your friend...in fact, sometimes it's useful to have more of it!

Why? For those without control room isolation (or recording themselves), if you monitor the signal with enough latency (e.g. a few seconds) you can completely seperate the sound you're evaluating with the direct signal also getting into your ears. You can also separate it from any visual cues that the end-listener won't be able to appreciate. This is useful for getting your tone printed right.

Anyway 128 samples of latency isn't the end of the world for many situations, but it is too long for recording vocals, where the direct signal in the singer's skull will have nasty phase interactions with that much latency. Low-latency monitoring in PTLE works fast enough on my machine (Mac Pro, 003) to cut vocals with.

Remember that you give up the ability to monitor with plugins active in a low-latency monitoring situation. You can use effects in the analog domain either going in or coming out to solve that. If you go purely in the analog domain for your cue mix, you won't be evaluating the effects of conversion (especially clipping!) if that's what you, the engineer, also are listening with.

Huh?
I started with Pro Tools about 6-7 years ago on a 001 and, became adaptive to the delay at 256. Then I progressed to the Digi 002 and used 64 and 128 and thought that was golden. Then switched to a mixer for monitoring and didnt really concern myself with the fact that there was zero latency UNTIL I did a session monitoring through the Digi 002 at 128 and it made it almost unuseable (distracting.) Studio Musicians can tell instantly while bedroom musicians are fine with it.

[Ethan Winer]

Originally Posted by nativeaudio

I thought that enabling software monitoring would enable something very useful that I would miss if I disabled that function. But is that really true?

I have a very fast computer but I never use software monitoring. Kiwi summed it up well:

Software monitoring sucks.

Here's how I do it:

Using a Mixer With a DAW

--Ethan

[tomdarude]

I ALWAYS monitor analog !!!!

(right about using any processing/verb on the cue-mix only with hardware)

this will make a HUGE difference in the musical outcome of a tracking session!!!
even if you can only barely hear the latency, the impact of it´s "feel" on the performer is way bigger than one might think! (e.g. on (acoustic)guitar, bass, vox, drums/perc)

for Overdubs my Dangerous MonitorST takes all this duties, one more thing why I´ll likely never ever regret this purchase!

cheers tom

[Geosync]

We are going to roll out the n Series mixers . In select stores in August. They will come with a special version of Cubase called Cubase AI

When used together, (or with Cubase 4) you can actually monitor wet (DAW) or dry (direct) on an input by input basis. I am really excited about this baby.
(It brought me out of hiding ;)

The Class-A Mic pres are rich and pristine and the new morphing sweetspot compression is pretty cool. Some of the presets were done by the Slutz own Michael Wagener and Chuck Ainlay.

Anyway, this low-latency monitoring will be a welcome addition to digital recording

[nativeaudio]

I agree that the low latency achieved by not using software monitoring is important, which is why I don't use software monitoring either. Are there any important Logic functions that needs software monitoring enabled to work properly? If not - why is software monitoring the default setting when direct monitoring is so much better?

Of course - without software monitoring, Logic's faders, mute buttons and pan knobs won't currently control the level and pan position of the dry signal, but that is because there isn't any protocol that allows Logic to control the dry signal (or there is a protocol, but it isn't used by Logic). I find this frustrating, not only because I don't want to be dependent on a hardware mixer, but because controlling the dry signal this way was possible using ASIO2's "Direct Monitoring" in OS 9 and Windows years ago. I also believe Digital Performer can do this today (at least with MoTu hardware?).

Originally Posted by Beardhead

For me it has always been out of question to use software monitoring. If you have to use effects on the cans, use outboard gear. It's a no go to influence the musical source in a bad way!

The only outboard gear I've kept is stuff like high-end preamps – no Lexicon reverbs in my rack anymore. Since I still can use software reverbs (although not as inserts) that isn't a real problem other than that it's a real pain to have to use workarounds to make reverbs on auxes work in real time.

Originally Posted by Geosync

When used together, (or with Cubase 4) you can actually monitor wet (DAW) or dry (direct) on an input by input basis. I am really excited about this baby.

What if you want to not only switch between direct and software monitoring pr. track, but also control pan and fader position for the dry signal? Can this be done from Cubase? That would be great, especially if this is done using a standard Logic also can use.

[Geosync]

Originally Posted by nativeaudio

I


What if you want to not only switch between direct and software monitoring pr. track, but also control pan and fader position for the dry signal? Can this be done from Cubase? That would be great, especially if this is done using a standard Logic also can use.

What the wet signal does is a function of the DAW. Cubase allows for panning the effect separate from the dry. In other words pan effect to the left and the dry right. Cubase 4 also has seperate surround panning for effects. I used it and it's pretty cool when you automate the effects moving in the surround space. Since the n Series is an "Analog like" mixer the pand movements are done with a mouse. I explain more about this in the new product announcement section.

[GuitarRuss]

Originally Posted by peeder

Sometimes latency can be your friend...in fact, sometimes it's useful to have more of it!

Why? For those without control room isolation (or recording themselves), if you monitor the signal with enough latency (e.g. a few seconds) you can completely seperate the sound you're evaluating with the direct signal also getting into your ears.

If I understand your point - Putting it at extreme latency while doing mic placement can help because when you are placing the mic you are hearing the combination of the source and the signal. So for example, you might not notice that the mic placement is losing a bit of bottom end because you're hearing it right from the guitar you are playing, which is mixing with the headphone sound. If you set the latency for a few seconds you will hear it totally independently without having to shuttle back and forth to your computer. After doing mic placement, you lower latency as much as possible for the actually performance or do direct monitoring for the performance, and keep latency down for mixing too.

I've never thought about doing this, but I think I will start as I do mic placement on my own guitars and have often been deceived by the blend effect so i have to shuttle back and forth, which is frustrating and wastes time. I imagine it's a bit distracting to hear the slapback, but I think I can handle that. Were there any other applications you use this for?

You can also separate it from any visual cues that the end-listener won't be able to appreciate.

I don't understand this point. What do you mean?

Some musicians are more sensitive to latency than others. I have to use direct monitoring because latency bugs me when recording guitar tracks. Most won't notice 256 or 128, but some will.

[dromedar]

I recently upgraded to a RME fireface 800 and now use the onboard digital mixer to set foldback mixes to the performers. I have found that drummers and vocalists often are most sensitive to even small amounts of latency (128 or even 64). No latency (or a very low latenc)y is crucial to a good performance.

[nativeaudio]

Originally Posted by Geosync

Cubase allows for panning the effect separate from the dry. In other words pan effect to the left and the dry right.

That's not something I normally need, but are you saying that you can change between monitoring the direct/wet signal from Cubase, in other words - is this:

a function that you don't need to leave Cubase to control? And, if this is the case, will you still be able to use the faders in Cubase to pan or adjust the level of the direct signal?

Originally Posted by dromedar

I have found that drummers and vocalists often are most sensitive to even small amounts of latency (128 or even 64). No latency (or a very low latency) is crucial to a good performance.

That's definitely my experience too, which is why I'm questioning if there's a good reason to make software monitoring the default, because software monitoring always means extra latency.

Maybe they have looked at a Symphony rig, @96k, with a minimum buffer setting (32), in an song not containing any softsynths, on the fastest Mac available, and concluded that software monitoring is usable for real work, forgetting that this is a setup that most likely less than 1% of the Logic user base is working on?

Life inside a testlab doesn't always reflect how things work in real life, and in this case (I've seen this example mentioned in the Symphony-ads) the test lab example - concluding that you can get 1.6 ms latency with software monitoring on - doesn't have much relevance for most Logic users.

[Geosync]

are you saying that you can change between monitoring the direct/wet signal from Cubase, a function that you don't need to leave Cubase to control?

Yes It's done with the n series, but yes you don't change screens or anything.

will you still be able to use the faders in Cubase to pan or adjust the level of the direct signal?

Of course, the n is not a control surface nothing changes with Cubase. You just have a better and more natural monitoring environment.

[nativeaudio]

Originally Posted by Kiwiburger

Software monitoring sucks. If your interface doesn't offer zero latency monitoring, use a mixer.

Originally Posted by Beardhead

For me it has always been out of question to use software monitoring.

Originally Posted by ricfoxx

Huh?
Then switched to a mixer for monitoring and didnt really concern myself with the fact that there was zero latency UNTIL I did a session monitoring through the Digi 002 at 128 and it made it almost unuseable (distracting.) Studio Musicians can tell instantly while bedroom musicians are fine with it.

Originally Posted by Ethan Winer

I have a very fast computer but I never use software monitoring. Kiwi summed it up well:

Originally Posted by tomdarude

I ALWAYS monitor analog !!!!

(right about using any processing/verb on the cue-mix only with hardware)

Originally Posted by dromedar

No latency (or a very low latenc)y is crucial to a good performance.

Since we all seem to agree that software monitoring is not the way to go, why, then, aren't Logic seriously optimized for working without Software Monitoring?

[Ethan Winer]

Originally Posted by nativeaudio

Since we all seem to agree that software monitoring is not the way to go, why, then, aren't Logic seriously optimized for working without Software Monitoring?

I don't use Logic, but I can't imagine it prevents you from monitoring the way I shown in my article linked above.

--Ethan

[Roy Harter]

What about using external mic-pre's into a DAW? What's the best way to monitor in THAT situation? It seems that latency would be inevitable. Even if you had a console. This is a problem that I've been having lately. I'm trying to bypass the console entirely.
Please correct me if I'm wrong...

[drummin4christ]

There are many interface's out there now that offer an onboard mixer that is direct monitoring. I use a Presonus Firestudio with a Digimax FS and can mix up to eight no-latency stereo headphone mixes. Life is beautiful now. If you need verb, you can route that thru these mixes and verb has latency already so a few more ms won't hurt.

[GuitarRuss]

Originally Posted by Roy Harter

What about using external mic-pre's into a DAW? What's the best way to monitor in THAT situation? It seems that latency would be inevitable. Even if you had a console. This is a problem that I've been having lately. I'm trying to bypass the console entirely.
Please correct me if I'm wrong...

I'm using a Mackie Onyx which has the pres and firewire card. On any decent mixer you have aux sends/returns which can be used to set up a headphone mix. If you are using standalone preamps and no mixer, you have to software monitor - which is why so many people have a mixer.

Thanks NativeAudio for the tip about using only the reverb from the DAW while keeping the monitoring direct from the mixer. The latency wouldn't matter on the reverb as many reverbs have a small delay before starting anyway.

[Ethan Winer]

Originally Posted by Roy Harter

What about using external mic-pre's into a DAW?

You can use Y splitters from the preamp outs to feed the DAW sound card and a small mixer at the same time.

--Ethan

[Roy Harter]

Thanks guys! Ethan, your real traps are brilliant. We use them in 3 of my rooms. And your article on monitoring with a console and DAW is very well written and informative. Kudos for taking the time to help!

[nativeaudio]

I made a poll about recording/monitoring here:
How is your recording/monitoring set up?