to go into DAW digitally, or not?

[famished]

Analog or digital, that is the question. Going in digitally means I don't have to invest in a pricey converter. The remaining question then is does anyone have a preference? I'm guessing those of you who don't have their DAW set for 48kHz will opt for analog in, but for those of you who are using 48kHz sessions, whaddaya think?

 

[randyman]

I routinely record at 48K/24 bit, as I really don't think 96K is worth chewing up the drive space for dense multitrack projects. (One-pass stereo for live location recording? Sometimes I do go ahead with the 96K.)

So, given 48K, I definitely go digital I/O. Why on earth add two passes of D/A and A/D conversion?

 

[AndrewSimon]

After having the same dilemma and doing all kind of tests
I concluded that the difference is negligible.
So I use analog.

 

[stevieray]

Maybe my needs are just different, but I think analog is the way I'm heading. I found digital to be way too much trouble even though I know it's the better choice latency-wise. Some of the things that frustrated me the most when trying to go digital were: 1) iTunes playback was wrong sample rate. 2) Output control was on a menu page (no real knob). 3) Having to change the sync setup (or having it change automatically) based on whether or not the Axe-Fx was on and connected.

Maybe there's some piece of the puzzle I'm missing, but I think not building in a wordclock connection and the ability to slave was a little short sighted. I'd hate to think I had a Big Ben in my setup and was forced to slave off the s/pdif out of my Axe-Fx just because it was the weakest link in the chain.

 

[theaero]

I use soundcards with awesome SRC (prism sound orpheus), so sample rate is a non-issue.

However, my fav part about going in digital is perfect recording levels, every time.

 

[kev]

I've yet to record with it but when I do I'm going to try both digital and analog. I've always put my guitar tracks through a Neve 1272 style preamp with a tube comp for more gooyness. So it'll be interesteing to hear the difference and whether the analog iron and glass adds any goodness to the sound or not, regardless of SRC.

 

[Rocket Brother]

I've yet to record with it but when I do I'm going to try both digital and analog. I've always put my guitar tracks through a Neve 1272 style preamp with a tube comp for more gooyness. So it'll be interesteing to hear the difference and whether the analog iron and glass adds any goodness to the sound or not, regardless of SRC.

If you find that the Neve and tube comp adds something you can´t get within the Axe or in your DAW, you could always try setting up a patch with them in the efx loop of the AXE, and then still go out digital - might be the best of both worlds and at least worth a try if you´re not already using the efx loop.
RB

 

[famished]

What I see is that there are conversions ahead no matter what I do. If I go in digitally, the conversion happens going from 48 to 44.1 for CD. If I go in analog, the conversion happens going D/A to A/D. Which is uglier?

 

[stevieray]

A/D D/A is uglier, and more importantly, slower. The latencies can add up quick.

 

[Scott Peterson]

From the Axe-FX (assuming use of a really good D/A converter) the difference is slight. Depends on the project, but for tracking (most setups I am working with use 44.1kHz (don't ask, not my setup!) or 96kHz) so I use the analog outs. If I am at home, I'll do it 48kHz and run digitally.

IMHO, YMMV.

 

[porieux]

Maybe there's some piece of the puzzle I'm missing, but I think not building in a wordclock connection and the ability to slave was a little short sighted. I'd hate to think I had a Big Ben in my setup and was forced to slave off the s/pdif out of my Axe-Fx just because it was the weakest link in the chain.

I have to agree with this unfortunately the biggest omission IMO. Wonder how much money it really saved?

 

[stvnscott]

I haven't done any recording of consequence with my AFX yet, but the noodling and practicing I've been doing through my DAW has been hooked digitally since I discovered the disparity in levels between the left and right analog outs. The digital solution gives me perfect levels every time.

I'm going into a MOTU HD192 and normally record at 24/48 anyway. The MOTU does a good job of SR matching even if I wasn't at 24/48. I can't hear the difference between digital and analog, but the latency is MUCH better going in digital.

One thing I do that is a littly wonky is I run an old, cheap ART Levelar tube compressor in the FX loop of the AFX. I have it set for almost no gain reduction. It is there for just a touch of color.

 

[Resonant Alien]

I run a Neve 1073-style preamp (Vintech x73i) and a Purple MC77 between the output of my Axe and the input to my DAW (Metric Halo ULN-2). I like the sound I get by running through the outboard gear versus going straight in digital. The difference is slight, and probably in a big mix, you would never notice it, but I think running through the analog gear gives the sound a more (...brace yourselves for totally overused and vague adjectives to follow....) open, airy, whatever, feel. I also run my guitar through a tube DI and into the rear input of the Axe instead of just plugging straight into the Axe, for the same reasons - again, it is a small difference, but I do notice a difference, and I like the sound of the pre-DI'd guitar better. YMMV. (Plus I feel better when I am using all my gear - makes me feel vindicated for all the money I've spent on it :mrgreen: )

 

[chase]

...since I discovered the disparity in levels between the left and right analog outs.

What disparity in levels?

If there's an unintended difference between the left and right analog output level, then you either have a hardware problem or a user error.

 

[VegaBaby]

One of the few but imho biggest drawbacks of the Axe is the 48 digital out. All sessions I do come in 44.1 or 96, never ever one of them in 48. My own stuff is in 44.1 as well so there's no way I can ever use the digital outs. In a way that really reminds me of those old Soundblaster cards people used to have problems with like there's no tomorrow...they were 48 only as well :lol:

 

[stvnscott]

...since I discovered the disparity in levels between the left and right analog outs.

What disparity in levels?

If there's an unintended difference between the left and right analog output level, then you either have a hardware problem or a user error.

All you need to know is contained here:

viewtopic.php?f=12&t=6083

Cliff confirmed he uses a dual-ganged pot to control the output for both L and R, which means there could be up to a 20% difference in output between them.

 

[VegaBaby]

...since I discovered the disparity in levels between the left and right analog outs.

What disparity in levels?

If there's an unintended difference between the left and right analog output level, then you either have a hardware problem or a user error.

All you need to know is contained here:

http://fractalaudio.com/forum/viewtopic.php?f=12&t=6083

Cliff confirmed he uses a dual-ganged pot to control the output for both L and R, which means there could be up to a 20% difference in output between them.

There were also some tets done which showed that low setting of the output control can lead to that. I think if one runs the output quite hot it's a lot less...

 

[stvnscott]

="VegaBabyThere were also some tets done which showed that low setting of the output control can lead to that. I think if one runs the output quite hot it's a lot less...

That depends entirely upon the specific pot in your AFX. In some cases, this will be true, but in other cases, it will be the opposite. You have to consider both the max value of each unit in the gang plus the taper of each unit in the gang. If you were unlucky enough to have a 0.8M (-20%) on the right and a 1.2M (+20%) on the left and the taper exactly perfect (which NEVER happens), you would have more disparity with the output dimed than you would at lower settings. But the taper of a carbon pot generally has more variance than the min/max values, so you have to include it in the equation. The taper of each unit in the example above could make the variance more or less at any given location on the dial--especially since we are talking about the notoriously inaccurate pseudo-logarithmic taper used in most commercial pots.

The "tets" you reference are purely anecdotal as the results will vary from one pot production lot to the next. In a few months, the same "tets" might yield the exact opposite results.

This is why most high-end stereo mic pres use rotary switches with 1% resistors to set output levels. This approach is less versatile in terms of continuous adjustment, but much more accurate in terms of level control from right to left.

 

[VegaBaby]

="VegaBabyThere were also some tets done which showed that low setting of the output control can lead to that. I think if one runs the output quite hot it's a lot less...

That depends entirely upon the specific pot in your AFX. In some cases, this will be true, but in other cases, it will be the opposite. You have to consider both the max value of each unit in the gang plus the taper of each unit in the gang. If you were unlucky enough to have a 0.8M (-20%) on the right and a 1.2M (+20%) on the left and the taper exactly perfect (which NEVER happens), you would have more disparity with the output dimed than you would at lower settings. But the taper of a carbon pot generally has more variance than the min/max values, so you have to include it in the equation. The taper of each unit in the example above could make the variance more or less at any given location on the dial--especially since we are talking about the notoriously inaccurate pseudo-logarithmic taper used in most commercial pots.

The "tets" you reference are purely anecdotal as the results will vary from one pot production lot to the next. In a few months, the same "tets" might yield the exact opposite results.

This is why most high-end stereo mic pres use rotary switches with 1% resistors to set output levels. This approach is less versatile in terms of continuous adjustment, but much more accurate in terms of level control from right to left.

Thanks for the technical explanation. Good to learn about that side, too.

Ahh...and thanks for pointing out my typo as often as possible . 'Note to myself':...Try not to leave out the letter 's' in the future...

 

[famished]

Ok, for those of you content to run 48kHz digitally, I have this to ask: Sure 48kHz is beyond the human hearing range in terms of frequencies that we can hear, but when I've read Bob Katz's book, he seems to come to the conclusion that 96kHz is still preferable, due to processing those sounds through plugins, etc. His idea, I think, is to start with 96kHz and bounce down from there. How do you guys feel about that?