Reamping adds major noise - what's wrong? [Solved]

Smittefar said:
I think cubase and most other DAWs has a function to measure the round trip latency with outboard gear and compensate automatically.
Click to expand...

I think that is true when the equipment is connected digitally. However, the AX8 can't reamp via USB, and so you have to do it in the analog realm. Cubase doesn't "know" how long it takes the signal to travel through the AX8, so I had to just manually.

If there is a way for cubase to "know" the roundtrip times in the analog realm, please let me know how to set this :). However, I think it only can do that when dealing with plugins and digitally connected equipment.
 
I am a little out of my waters here, but I think, you can ask Cubase to send a pulse out of your output and measure,when it shows up at the input. It's doing the exact thing you did, it is just automated.

I don't know Cubase, but Reaper has a plugin called ReaInsert that does nothing except sending sound out thru a designated output and listen to a designated input. It also tries to compensate for and measure latency.
 
REAPER at least has the ReaInsert plug which can detect delay on hardware outboard gear. I'm going to just try that and then use the transient detector on a track and reamped track to see if it's right down to a sample :)
 
Maybe someone could explain why the impedance mismatch would introduce noise. I understand that a high z source would suffer from a low z sink which is why a DI box is needed between a passive PU and a mixer input. But a low z source doesn't care if it is loaded by a high or low z sink, it will be able to provide the same signal level to both. If the source signal itself carries a lot of noise then of course the amp gain will make it audible. But if I'm not mistaken then the source of the noise is the AD/DA converters plus OpAmps along the signal chain which is where pro h/W makes the difference.
 
I believe the noise is proportional to the square root of the input impedance - It's the inherent noise in an impedance.

EDIT: Scratch that - That does not make any sense. I don't know, why it should be better to run a high impedance signal for re-amping unless you are re-amping through fuzz pedals or the like that need to load the pickups.
 
Last edited:
naka2112 said:
I think that is true when the equipment is connected digitally. However, the AX8 can't reamp via USB, and so you have to do it in the analog realm. Cubase doesn't "know" how long it takes the signal to travel through the AX8, so I had to just manually.

If there is a way for cubase to "know" the roundtrip times in the analog realm, please let me know how to set this :). However, I think it only can do that when dealing with plugins and digitally connected equipment.
Click to expand...
Not so. Using hard as an external insert you can have cubase or any DAW ping round trip latency.
 
ChristThePhone said:
Maybe someone could explain why the impedance mismatch would introduce noise. I understand that a high z source would suffer from a low z sink which is why a DI box is needed between a passive PU and a mixer input. But a low z source doesn't care if it is loaded by a high or low z sink, it will be able to provide the same signal level to both. If the source signal itself carries a lot of noise then of course the amp gain will make it audible. But if I'm not mistaken then the source of the noise is the AD/DA converters plus OpAmps along the signal chain which is where pro h/W makes the difference.
Click to expand...
I'm fairly certain that it isn't. What does happen though is that your instrument level signal in the process of being recorded is gonna be line level. Way hotter...
On the way out it'll by necessity increase noise because well it's a hotter signal hitting the input of whatever device you're re-amping.

The impedance mismatch comes into play in tend of top and bottom end being strong when manually stepping the level down.

Quick fixes without spending money on a re-amp box.
Using a buffered pedal such as a biss OD in off state will be your impedance quick fix on the way out to re-amp.
Flipping a DI and using it in reverse will step down the level.
It's not 100℅ but will get the job done.

Now, as for phase... Or as you guys think of it latency.
When printing one pass fine if you manually adjust for latency.
However the second you change blocks inside the Fractal your latency changes by roughly 2 msec...this is enough to mess with phase.

Instead of worrying about new models and more parameters to be put in the box... Use two cabs and adjust the delay time on one while listening to the resulting comb filtering of the eq.
Do this hard panned... When the signal goes to center you're phase coherent.
 
kvebbs said:
REAPER at least has the ReaInsert plug which can detect delay on hardware outboard gear. I'm going to just try that and then use the transient detector on a track and reamped track to see if it's right down to a sample :)
Click to expand...
You need to do this with clean sound since the wave form on overdriven it worst high gain sounds starts before the note from the sound initial noise.
 
I tested Focusrite 18i8 2nd Gen, Focusrite ISA One and AX8 as DI boxes today, and the difference is minimal

I recorded the same piece with the same electric guitar through each DI. An instrument input on the interface, ISA One going to a line input on the interface and AX8 going to a line input on the interface. I tried to match the recording levels to the best of my ability, so all three tracks peak between -4.1 and -5.4 dB, and then I measured the noise. If I turn down the volume knob on the guitar (effectively grounding the input) or if I have open circuits (nothing connected at the inputs), all DI boxes measure a noise floor of -91 dB, so I expect that is the noise floor of the interface. With the guitar connected, the volume opened up while dampening the strings (a very unscientific method), I measured -76 dB, -80 dB, and -81 dB on the interface, the ISA, and the AX8, respectively.

My conclusion is that you cannot claim that the AX8 is any worse as a DI than a regular DI.
 
A DI signal peaking at -4dB is stupid hot for re-amping. I'd expect you to have crazy leveled of noise. I'm not even sure how you get these levels unless you increase input gain.
You want -18dBFS... Pretty much your wave form will look like next to nothing is happening.

And how do you compare the signals? The Ax8 has to go into an interface right?
On an Axe2 I get that you can compare since it's a sound card.
So your not comparing DI to ax8 you comparing (presumably) instrument in on a sound card to Ax8 used as DI.
 
I would normally never use -4 dBFS, but the ISA One could not go any lower, so I chose that level to compare apples to apples.

The three different chains are

1) Instrument input on interface

2) DI input on ISA One -> ISA One Line Out -> Line in on Interface

3) Input 1 on AX8 -> Vol Block with +12 dB level -> AX8 Output 1 -> Line in on interface

(Line in on the interface is straight to A/D, no mic pre or gain adjustment in between)

(It does not hurt to record hot levels, as long as you don't go over. I would rather record a hot DI track and turn it down before re-amping than record a DI track that is too low and suffers from noise. When I reduce the volume of the DI track for re-amping, I also reduce the recorded noise level. That said, I normally aim to record with a peak level around -12 dBFS to make sure there are no overs.)
 
Not sure why you'd put a volume block at +12.

Ok... So recording DI hit and turning down... Bringing down the noise floor...
In order to print a track used for reamping that is printed at hot leveled you already amplify the noise. So if you don't print hot there's less noise to content with.

As for levels and noise floor I have yet to meet guys engineering stuff with consumer or producer gear that have a solid handle on gain staging or overhead.
The tracks I see are always hit enough that I have a template that pads everything 12dB.
 
I put a +12 volume block in to make sure, I record all three at the same level.

Whether the noise is amplified, when printing hot depends on the origin of the noise. If the origin is the A/D converters, hitting the A/D converters with a hot (but not over) level gives you better signal to noise ratio. If the noise is coming from some electronics before the A/D, SNR might be the same.

I just discovered this video



@ChristThePhone Looks like the re-amp box is mostly about reducing the level in the analog world rather than the digital world, which makes sense :) They don't seem to be concerned about impedance.
 
I'm not sure why were even have this discussion. Gear is designed to run at -18dbFS, so if you print cause you subscribe to the galaxy if using "all the bits" what you gonna get is overcooked meat, if you they're it in the freezer you can chill it down but it's still overcooked out burned. If you leave it medium rare you can still heat it.
Besides having spent my live on both sides of the glass for fit 4+decades. I also started, overlooked and taught a production course at a music academy.
Your obviously welcome to print your tracks any which way you chose but your not going to get me to agree that hitter makes any sense.
 
Smittefar said:
I put a +12 volume block in to make sure, I record all three at the same level.

Whether the noise is amplified, when printing hot depends on the origin of the noise. If the origin is the A/D converters, hitting the A/D converters with a hot (but not over) level gives you better signal to noise ratio. If the noise is coming from some electronics before the A/D, SNR might be the same.

I just discovered this video



@ChristThePhone Looks like the re-amp box is mostly about reducing the level in the analog world rather than the digital world, which makes sense :) They don't seem to be concerned about impedance.
Click to expand...

That looks like radial jDI not xamp. Didn't watch.
The camp has Hi-z out to deal with the impedance from line to instrument. And can take 22 dBU.
Seeing that nominal operating level of audio gear is assumed to be 0dBU ie.-18dBFS that's for guys printing insane levels at up to O dBFS.
 
Ed, did you miss the post where I wrote that I printed -4 dBFS out of necessity, and it is not what I usually do? My point is only that it does not change the validity of the test.

He tests both the JDI and the x-amp.
 
You must log in or register to reply here.
Back
Top Bottom