Jump to content

external hardware,null test


Recommended Posts

Hi

 

I have a saffire pro 40 interface that I just got to essentially utilise my outboard harware with (EQ and COMP).

 

Up until now for the last year I have been using an apogee duet.

 

I am using Logic 9.(everything up to date including OSX)

 

To test things out I tried a little test.

 

I have an acoustic guitar part recorded.

 

I inserted Logics I/O plug in.

 

I have a balanced cable running from the saffire pro 40's number 4 output to its number 2 input.

 

I ping'd the I/O plug in in Logic and got 25 samples...all good.

 

Now the acoustic guitar is routed out of logic and back into the same channel.

 

I made a copy of the acoustic guitar track, and then put it out of phase (this of course doesnt have the I/O plug).

 

In theory, you should get a perfect null.

 

Of course reality, is you will hear the AD converter and any level differences.

 

My system runs at 0VU = -20dBfs.

 

The closest I can come to nulling, is around -40db with my guitar signal going out and back into logic at my 0VU point.

 

I was expecting to be much closer than that.

 

I have been mucking about with this for an hour or so.

 

How far down, would you expect the null to be, I would have thought really really low..at least 80db down.

 

I can of course remove the I/O plug in from the first track and then I get a perfect null.

 

So, to summarise, the lowest I can get the null to be, running the saffire pro's output to an input, is around 40dB down.

 

any thoughts.

 

cheers

 

Wiz

Link to comment
Share on other sites

I ping'd the I/O plug in in Logic and got 25 samples...all good.

Did you enter "-25" for the Latency Offset parameter?

 

not quite sure what you mean.

 

I ping'd and the latency offset was set to 25 samples...so I figured that its compensated....it automatically sets itself, when you ping.

 

cheers

 

Wiz

Link to comment
Share on other sites

it automatically sets itself, when you ping.

Not here.

I think it's supposed to...but it never has worked here.

 

Set the Latency Offset to 0.

Now, watch the Latency Offset slider when you click the Ping button.

Did it move to the right as it reads "+25"?

 

If it automatically set itself, it should move to the left and read "-25".

 

I bet if you enter "-25" for the Latency Offset, it'll work correctly.

 

...Unless your signal goes out of Logic, out of your interface, through your comp and EQ, back into your interface, and back into Logic and gets there 25 samples TOO EARLY. :shock:

 

The above scenario can happen.

Depends on the drivers for the interface.

 

Were you running through your outboard comp and EQ when you did the test?

Link to comment
Share on other sites

it automatically sets itself, when you ping.

Not here.

I think it's supposed to...but it never has worked here.

 

Set the Latency Offset to 0.

Now, watch the Latency Offset slider when you click the Ping button.

Did it move to the right as it reads "+25"?

 

If it automatically set itself, it should move to the left and read "-25".

 

I bet if you enter "-25" for the Latency Offset, it'll work correctly.

 

...Unless your signal goes out of Logic, out of your interface, through your comp and EQ, back into your interface, and back into Logic and gets there 25 samples TOO EARLY. :shock:

 

The above scenario can happen.

Depends on the drivers for the interface.

 

Were you running through your outboard comp and EQ when you did the test?

 

 

Well this is curious.

 

Were you running through your outboard comp and EQ when you did the test?

 

Nope, I have the output directly connected to the input.

 

Set the Latency Offset to 0.

Now, watch the Latency Offset slider when you click the Ping button.

Did it move to the right as it reads "+25"?

 

 

I did just as you described, I set latency offset to 0, I clicked ping, slider jumps to right, and reads +25 samples.

 

Still sounds like it did before, not quite nulling, down around -40dBfs

 

If it automatically set itself, it should move to the left and read "-25".

 

I bet if you enter "-25" for the Latency Offset, it'll work correctly.

 

Well, I manually adjusted the slider to -25 samples, and it got much louder, as I slowly go across to the right, stopping logic each time, adjust latency detection, then play ....it gets softer until +25 samples is the softest it will get, I continue towards the right , passed +25 samples and it starts to get louder.

 

 

Are you sure, that the latency detection figure, is meant to read negative? I assumed it was adding that figure to the rest of the tracks in the project.

 

thanks for your help with this.

 

 

What sort of null cancellation can you get on your interface...and are you sure it reads "-" on your latency detection and not "+"?

 

thanks

 

Wiz

Link to comment
Share on other sites

the plot thickens...8)

 

 

Okay I opened a new projet.

 

 

I have no plug ins inserted at all.

 

I have a single tick on each beat at 120bpm in an audio file...I found on this site somewhere.

 

I drag it in to a audio track. It has a single tick, every bar. Beauty!

 

I conneted output 4 of my interface to input 2.

 

Routed the output of the tick track, to output 4.

 

Created a new track and set its input to 2.

 

recorded the tick file.

 

Using the sample editor, it is indeed 25 samples late.

 

So, I set the record offset delay in Preferences/Audio to -25 samples.

 

Rerecord the tick file.

 

BINGO! perfect alignment, sample accurate.

 

 

Great.

 

So now I insert the I/O plug in on the tick track.

 

set its output and input up as it needs, ping it.. +25 samples.

 

Record the tick track....its wrong.

 

 

So I set the I/O plug off set to 0.

 

Record it . its wrong its actually hundreds of samples early.

 

 

maybe somewhere in the order of 700.

 

 

So basically. The I/O plug in is not working for me.

 

I dont know whats going on.

 

But, by using record offset, I can get sample accurate so that maybe what I have to do, and just set up routing.

 

weird stuff

 

Wiz

Link to comment
Share on other sites

Wizo,

Setting record delay vs setting the sample offset in the I/O plug are very different things, for very different work scenarios. It's a good idea to set up your record delay first, then leave it alone.

 

I'm actually surprised that you're getting consistency in pings from the Safire. Since it uses native firewire drivers and doesn't have its own. Note that that stability could change if you change the demands/devices on the FW bus.

 

If you tell us how you want to actually use the external path in your mix, we can suggest an appropriate routing and patching.

 

-40 null sounds about right for comparing a digital to an analog path using mediocre converters, especially at a 44.1 sample rate.

Link to comment
Share on other sites

Wizo,

Setting record delay vs setting the sample offset in the I/O plug are very different things, for very different work scenarios. It's a good idea to set up your record delay first, then leave it alone.

 

I'm actually surprised that you're getting consistency in pings from the Safire. Since it uses native firewire drivers and doesn't have its own. Note that that stability could change if you change the demands/devices on the FW bus.

 

If you tell us how you want to actually use the external path in your mix, we can suggest an appropriate routing and patching.

 

-40 null sounds about right for comparing a digital to an analog path using mediocre converters, especially at a 44.1 sample rate.

 

Hey thanks for the reply and the offer of help.

 

Okay...well I just want to incorporate outboard hardware. Comps EQ etc.

 

I want to be able to route audio tracks , out of logic, essentially reamp them through the EQ and or COMP, and re record them into logic.

 

I wanted to use the I/O plug to set up the outboard EQ and COMP, and then print a copy of the reamped track to a new track so I could mute and hide the original, so it was there to be reamped later.

 

I calculated my recording delay, which is 25 samples and is repeatable.

 

If you think about it, the I/O plug in delay and the recording offset delay in this case should be identical...at 25 samples...because all I have is a connection between the saffires output and input.

 

Also that delay shouldnt change, if I never change my project settings, which I dont. Always same buffer, sample rate and bit depth. (you mentioned sample rate effecting the quality of the null achieved...I wonder how? How would sample rate have any effect on how well a singal nulls?)

 

Also, it shouldnt change no matter how many pieces of analog outboard I use, as well...they are analog...8)

 

So I think I have a pretty good handle on this. But, as soon as I implement the I/O plug into the path...all hell breaks loose....8).

 

So, is for everyone else, the I/O plug working flawlessly....??

 

When you wire up and output to an input, and ping, does the I/O plug report the same delay as manually calculating your delay?

 

How do you guys print your tracks when using the I/O plug? (THIS IS THE BIG QUESTION...8)....)

 

thanks

 

Wiz

Link to comment
Share on other sites

Wizo,

Setting record delay vs setting the sample offset in the I/O plug are very different things, for very different work scenarios. It's a good idea to set up your record delay first, then leave it alone.

 

I'm actually surprised that you're getting consistency in pings from the Safire. Since it uses native firewire drivers and doesn't have its own. Note that that stability could change if you change the demands/devices on the FW bus.

 

If you tell us how you want to actually use the external path in your mix, we can suggest an appropriate routing and patching.

 

-40 null sounds about right for comparing a digital to an analog path using mediocre converters, especially at a 44.1 sample rate.

 

Hey thanks for the reply and the offer of help.

 

Okay...well I just want to incorporate outboard hardware. Comps EQ etc.

 

I want to be able to route audio tracks , out of logic, essentially reamp them through the EQ and or COMP, and re record them into logic.

 

I wanted to use the I/O plug to set up the outboard EQ and COMP, and then print a copy of the reamped track to a new track so I could mute and hide the original, so it was there to be reamped later.

 

I calculated my recording delay, which is 25 samples and is repeatable.

 

If you think about it, the I/O plug in delay and the recording offset delay in this case should be identical...at 25 samples...because all I have is a connection between the saffires output and input.

 

Also that delay shouldnt change, if I never change my project settings, which I dont. Always same buffer, sample rate and bit depth. (you mentioned sample rate effecting the quality of the null achieved...I wonder how? How would sample rate have any effect on how well a singal nulls?)

 

Also, it shouldnt change no matter how many pieces of analog outboard I use, as well...they are analog...8)

 

So I think I have a pretty good handle on this. But, as soon as I implement the I/O plug into the path...all hell breaks loose....8).

 

So, is for everyone else, the I/O plug working flawlessly....??

 

When you wire up and output to an input, and ping, does the I/O plug report the same delay as manually calculating your delay?

 

How do you guys print your tracks when using the I/O plug? (THIS IS THE BIG QUESTION...8)....)

 

thanks

 

Wiz

 

Well I think I just solved it...8).

 

It may very well be a routing issue.

 

I set up a pre fader send to a bus, on the channel that the I/O plug is inserted in.

 

I set that send level to unity.

 

I created a new track and set its input to the output of the bus.

 

Fantastic, it recorded , sample accurate.

 

 

This is not so bad....so what I will probably do, is always use the same buss number, say 99. And it will sit in the project and I will use it to print whatever is sitting connected to my interface at whatever time....

 

 

cool....

 

 

thanks everyone for sparking me along.

 

cheers

 

Wiz

Link to comment
Share on other sites

Also that delay shouldnt change, if I never change my project settings, which I dont. Always same buffer, sample rate and bit depth. (you mentioned sample rate effecting the quality of the null achieved...I wonder how? How would sample rate have any effect on how well a singal nulls?)

Converters have to filter the sample rate. The lower the SR, the closer that filter's cutoff is to the audio signal. Which means more artifacts and more phase shift/group delay.

 

Fantastic, it recorded , sample accurate.

Good to hear you got it going. But from your description are you sure you're recording from the analog input? The bus only contains what's being sent "to" the aux, not what's being returned.

 

Alternatively, bounce your track in real time with the channel solo'd. You will need to solo-safe the output channel that sends to your effects. Grab the new region from the bin and drag to a new track. Use the Move to Original Record Position command.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...