Recabinet 2.0 Modern and Vintage - Out Now!

[javajunkie]

Funny, I liked the "Top Boost 2x12" the most...
Pair it with "Class A" amp and there is some magic going on there.
I had high hopes for the Fender cabs but they sound too bright/thin to me.

I guess it's a matter of taste, anyway I'm sure I will find at least a few useful ones and that's worth the price for me.



PS the "Top Boost 2x12" at 45 degree angle is the one I'm talking about.

Well there you go, I hate off axis . I just have a ones I got off guitar rig and modded that I like better. Oh well.

 

[fremen]

I tested half of the modern cabs from the demo version and they don't do it for me, they lack low end. Even cabs from the 1.05 version had more low end - or is it because I tweaked those cabs :?

 

[Robboman]

as well as FULL AxeFx support this time, complete with reference mic "far field" IRs of each cabinet.

I'd be very interested in a description of all the conditions under which farfield IRs were acquired, including speaker and test mic locations relative to (and distances from) room boundaries, and the distance from test mic to the speaker.

FWIW, I would not characterize a $99 microphone sold for use with an inexpensive measurement system as a "reference mic."

It never even occurred to me, I have one of those mics, bought it with my DBX Driverack a few years ago to RTA / EQ my band's PA. Pretty sure it's not Earthworks calibre, but I sure would like to capture usable IRs of my own cabs, because one of them is a gem, IMO, (1980 Marshall 4x12 w\ orig G12H-65s) but it pretty much just lives in my garage. Is it even worth the bother with this DBX mic?

(Sorry if this counts as a thread hijack)

 

[GuitarDojo]

I would like to do an IR of my Laney 4x12 H&H w/ vintage 60 speakers.

 

[BradMc]

Well, I just went through the Top Boost 2x12. Unfortunately the seems to be so colored by the amp they haven't worked well at all for me on the axe-fx.

I have a feeling it is going to be that way for all of them that were captured through the preamp.

I went through the vintage ones and the tangerine (the ones I intended to use). They only one that will work for me is green 4x12 w/ capture w/ 6l6 poweramp.

Oh well, at least I don't have to worry about removing the leading silence out on all of them.

I`m finding the only one that works for me so far is the Green 4x12 also, but it`s got a less low end than the stock Axe IR`s.

Using the Reference mic IR`s and adding the R121 from the Axe helps, i guess i`ve gotta give it more time to tweak these to work with the Axe.

 

[javajunkie]

Well, I just went through the Top Boost 2x12. Unfortunately the seems to be so colored by the amp they haven't worked well at all for me on the axe-fx.

I have a feeling it is going to be that way for all of them that were captured through the preamp.

I went through the vintage ones and the tangerine (the ones I intended to use). They only one that will work for me is green 4x12 w/ capture w/ 6l6 poweramp.

Oh well, at least I don't have to worry about removing the leading silence out on all of them.

I`m finding the only one that works for me so far is the Green 4x12 also, but it`s got a less low end than the stock Axe IR`s.

Using the Reference mic IR`s and adding the R121 from the Axe helps, i guess i`ve gotta give it more time to tweak these to work with the Axe.

That is precisely what I noticed and Cliff's original cab are not what you call bassy to start with.

 

[GuitarDojo]

Any updates Kazrog?

 

[Kazrog]

Any updates Kazrog?

We're pretty busy lately, but we have been conducting some research into this matter. So far my initial impression is that the IRs don't sound as "real" without the leading audio before the spike. It's a subtle thing, but it is audible and there are noticeable differences when viewing the spectrum with the meter in Cubase 5 with aligned, processed audio using trimmed and non-trimmed IRs (in particular, the audio processed with non-trimmed IRs has more content at the extreme high and low end of the spectrum.)

Obviously, even a very small amount of latency like we're discussing is best avoided, but not at the expense of tone. We're figuring out if there's a way to reduce the latency without compromising tone, and I'll post here again once we try some more things in this regard.

 

[Jay Mitchell]

So far my initial impression is that the IRs don't sound as "real" without the leading audio before the spike.

Your impression is wrong. No real speaker delays the signal by 7.5ms, nor does any real speaker need 7.5ms added delay to produce "good tone."

It's a subtle thing,

7.5ms added latency is anything but subtle, and it hurts, rather than helps, "tone" (more properly, the accuracy of representation of the speaker's response ), when you do not have the option of convolving with an arbitrarily long impulse response. It is very common for deconvolution algorithms not to preserve absolute time, and yours is obviously among the ones that do not. The initial "silence" is not the result of anything that occurs physically, it is artificially generated by the deconvolution algorithm. Removing it compromises nothing and improves everything of importance here.

(in particular, the audio processed with non-trimmed IRs has more content at the extreme high and low end of the spectrum.)

No. An IR with fewer data points (the effect your leading silence is causing) has less detail at all frequencies. In order to do an accurate analysis of an IR, you must limit the spectral conversion (FFT) to the total number of samples that will actually be used in the target IR. In the case of the Axe-Fx, the number is 1024 points in "Hi-res" mode and 512 points in "Lo-Res" mode, and the IR will begin with the first data sample if you use AlbertA's free converter. When you've artifically added 7.5ms of leading silence, "Hi-Res" mode becomes 667 points and "Lo-Res" is 152 points. We had numerous discussions on the old board about the audibility of the difference between 512 and 1024 points (which we had to do "creatively" prior to Cliff's giving us the "Hi-res" option). Everyone who made the comparison agreed that 1024 points sounded noticeably better. What do you figure are the odds that these people will perceive 152-point IRs as acceptable? That's what you're offering them now.

Obviously, even a very small amount of latency like we're discussing is best avoided, but not at the expense of tone.

If eliminating it comes at the "expense of tone," then you're doing something badly wrong.

We're figuring out if there's a way to reduce the latency without compromising tone,

Yes, there is. Edit out leading silence, and thus shift the entire impulse response backward in time, where it resides in the physical world. The only effect on spectral content will be due to the added data you now have within the IR window, but that is a good thing.

 

[DieSchmalle]

I totally agree with Jay here.

The leading silence in those Recabinet .syx files show either lack of knowlege or lack of passion of the one who created them.

Would be ok if it was a nonprofit thing. Resolution is wasted and latency added. Not professional.

Nonetheless I like the Recabinet IR's soundwise.

DieSchmalle

 

[javajunkie]

Any updates Kazrog?

We're pretty busy lately, but we have been conducting some research into this matter. So far my initial impression is that the IRs don't sound as "real" without the leading audio before the spike. It's a subtle thing, but it is audible and there are noticeable differences when viewing the spectrum with the meter in Cubase 5 with aligned, processed audio using trimmed and non-trimmed IRs (in particular, the audio processed with non-trimmed IRs has more content at the extreme high and low end of the spectrum.)

Obviously, even a very small amount of latency like we're discussing is best avoided, but not at the expense of tone. We're figuring out if there's a way to reduce the latency without compromising tone, and I'll post here again once we try some more things in this regard.

I don't get the logic behind leading silence altering the tone (other than truncating the end of the IR in the axe-fx).

 

[Scott Peterson]

I bought the upgrade, but haven't had any time to sit down with them to even give them a listen.

Reading all this here just goes to show that you can sometimes learn a hell of a lot by just reading and knowing whom to trust.

Frankly, if there is 7.5ms of latency dropped into my rig, I'm not a happy guy. That's what I used to suffer through to record and monitor back almost 9 years ago. In this day and age, not acceptable.

I am hoping that an update comes from Recabinet that follows some of the expert opinions here and there is no latency or leading silence knocking down the depth of the IR bit rate by wasting a huge chuck of it on silence.

Silence is golden, but not when it comes to IR's.

 

[Matman]

it is audible

Could this be because of a bug or "feature" in your convolution playback system?

 

[BradMc]

it is audible

I dont actually hear any delay when using the Re-Cabinet IR`s through the axe.

 

[Jay Mitchell]

What follows is IMO. Take it for what it's worth.

It's not exactly a confidence-builder to discover such a substantial oversight in the processes employed to generate data being offered for commercial sale. It is quite clear that either the data was not closely examined, or the person doing the examination did not have a sufficient grasp of the relevant issues to be able recognize that a problem exists.

Having been kind enough to point out the problem, had it simply been corrected, I could chalk all this up to excessive haste in bringing a product to market. It is positively jaw-dropping to see the purveyor of the data defending this defect as if such a serious "bug" could possible be a "feature." I had my doubts before, but they have now been completely removed.

 

[Kazrog]

Jay - I very much appreciate you taking the time to leave a detailed response. What you're saying with respect to IR resolution is absolutely true with respect to the AxeFX, but it is not true for applications such as SIR2, Logic Space Designer, Altiverb, etc. because they don't artificially chop the IR after a certain number of samples.

And yes, it's also true that no guitar speaker setup is going to introduce 7.5ms of latency, and that the vast portion of that content is junk introduced in deconvolution, however, the key to getting this right is finding the exact point at which "junk" ends and "necessary" begins. Or taking another approach to deconvolution that doesn't introduce latency to begin with. This is the heart of the matter, and is why we're putting research into it rather than just doing a quick batch chop.

Recabinet supports a lot of different environments, the AxeFX being one among many, and I don't want anyone to have a lesser experience. I realize that right now, AxeFX users are getting exactly that, which will be fixed soon, I just want to make sure that the update patch is done right.

Again, thank you for your time.

PS - BradMc - nobody is going to "hear" 7.5ms of latency on its own, but it can create some very audible phase issues if you run the identical guitar signal through something else with zero latency in parallel.

 

[javajunkie]

the key to getting this right is finding the exact point at which "junk" ends and "necessary" begins

Junk ends? It is silence?

 

[Kazrog]

What follows is IMO. Take it for what it's worth.

It's not exactly a confidence-builder to discover such a substantial oversight in the processes employed to generate data being offered for commercial sale. It is quite clear that either the data was not closely examined, or the person doing the examination did not have a sufficient grasp of the relevant issues to be able recognize that a problem exists.

Having been kind enough to point out the problem, had it simply been corrected, I could chalk all this up to excessive haste in bringing a product to market. It is positively jaw-dropping to see the purveyor of the data defending this defect as if such a serious "bug" could possible be a "feature." I had my doubts before, but they have now been completely removed.

I can understand how you've arrived at this perspective, and given the sum total of what's been said, I don't hold this perspective against you. However, you're not getting where I'm coming from with this, and I'm sorry if I was unclear at all.

To clarify, my post was certainly not intended to "defend" anything, it was to provide an initial observation based upon actual research (because I was asked for an update, and this is where things stand at the moment.) That observation was also presented with the understanding that we are continuing to look into the matter and arrive at the best solution, not just the first or easiest solution.

Oversights, bugs, etc. occur with any software or data product. A lot of time and hard work was put into the 2.0 release by a small team, and somehow this made it through to release anyway. Yes, it is a significant oversight, but yes, we are going to fix it. This is a fully supported product, and we are putting our focus into getting a fixed update out as soon as possible.

 

[Jay Mitchell]

What you're saying with respect to IR resolution is absolutely true with respect to the AxeFX, but it is not true for applications such as SIR2, Logic Space Designer, Altiverb, etc. because they don't artificially chop the IR after a certain number of samples.

Every system that can do real-time convolution has a definite limit on the length of the IR used for convolution. Even non-real-time systems have practical limits. The IRs of yours I've looked at contain ~8640 samples. They would be compatible with, at most, an 8192-point convolution. That means that the IR is "artifically chopped" at that number of samples.

Furthermore, any application that employs convolution in a multitrack recording environment will preserve the ~7.5ms latency added by your IR. If you were, for example, to mix two guitar tracks, one direct and the other convolved with your IR, at similar levels, the 7.5ms latency in the IR would cause deep frequency response cancellations at 67Hz, 200 Hz, 333Hz, 467Hz, and all other odd integer multiples of 67Hz. If you want this comb filtering effect, you could easily add the required 7.5ms delay to one of the tracks. OTOH, if you're unaware that your "cab" IR is delaying the track with which it is being convolved, you could easily wind up scratching your head wondering why the sound got so weird all of a sudden. Ergo, removal of leading silence is a positive thing for any application, not just the Axe-Fx.

And yes, it's also true that no guitar speaker setup is going to introduce 7.5ms of latency, and that the vast portion of that content is junk introduced in deconvolution, however, the key to getting this right is finding the exact point at which "junk" ends and "necessary" begins.

That's where knowing something about spectral transforms and loudspeaker behavior can come in handy. It's helpful, for example, to know how precise you need to be in identifying the initial point for the IR. You don't need (and will never be able) to narrow it down to a single sample.

 

[Kazrog]

Jay - rest assured, nothing you say in the above post is news to me or anyone else on the Recabinet team.

I realize that limits do exist in every convolution host, but none of them we support have as low of a limit as the AxeFX, so that is why the resolution issue caused by the leading silence exists specifically for AxeFX users. I also realize that the latency affects everyone, and I understand that comb filtering arises when audio processed through the current Recabinet 2.0 IRs is run parallel with the same signal processed by another IR that lacks latency, in fact I already addressed that in a previous post in this thread.

Bottom line: a mistake made its way into our 2.0 release, and we will release a free update to fix it very soon. I apologize for any inconvenience this has caused for anyone.