Reverb comparison

AlbertA

AlbertA

Fractal Fanatic
So here's a nice little stress test dry clip for reverbs:

http://www.numericalsound.com/RI_ShootOut/TitBrSnare.mp3

Here's a recording of all the reverb presets using their defaults, 100% WET.
http://guitarlogic.org/axe-fx2/reverb/reverb_axefx_test.mp3

The Axe-Fx does fares well in that there's no obvious artifacts and it's absent of obvious metallic qualities - there's a very slight metallic quality but it fares better than other reverb simulators using allpass networks.
It does lack a bit of echo density though compared to new simulator plugins, which I think its most evident on the small to medium spaces.

To show what I mean I also ran the same test loop through ValhallaRoom - which is a new reverb simulator plugin released last year which should be a fair comparison as there are they are both simulators not convolution reverbs. To my ears it does a wonderful job in the density department. It excels in the small space department too, which clearly outperforms the Axe-Fx reverb IMHO. The stereo quality is also more evident that the Axe-Fx reverb. In my core i7 One instance of it hardly uses 1% of the cpu (I know block sizes will be different, but it shows you how much can be done).

These is just a few of the built in presets of ValhallaRoom also 100% WET:
http://guitarlogic.org/axe-fx2/reverb/valhalla_test.mp3

The test clip came from this shootout page, where you can also hear the difference when using a convolution reverb (Although that would be an unfair comparison).
FORTI and SERTI Convolution Libraries from Numerical Sound
http://www.numericalsound.com/RI_ShootOut/TitBrSnare_SERTI_Tail30.mp3
http://www.numericalsound.com/RI_ShootOut/TitBrSnare_FORTI_Epic1.mp3

That's not to say the Axe-Fx reverb is unusable, not by a longshot....However, given the great advances in amp modeling, I was expecting a bit more in the reverb department (I know there was no indication of this, just wishful thinking :).

My wish is for an enhanced reverb algorithm in the vein of ValhallaRoom. If not in Axe-FX 2, then in whatever future product you may or may not be working on :)
 
Last edited:
I would agree that the AFXII reverb is not totally up to current sonic standards, but I feel sure there will soon be an update, since reverb is 'the mother of all effects'...
 
Reverb test?
"Holy Grail" test, IMO, is, and always will be, well recorded vocals.
If a 'Verb can stand up to scrutiny when used on a great vocal track... :D
 
AlbertA said:
\The stereo quality is also more evident that the Axe-Fx reverb.
Click to expand...

The axe reverb is a mono to stereo type, which does not give a realistic stereo image. The Valhalla is true stereo (which is actually quadraphonic - note that this differs from parallel stereo), and gives a far more realistic sense of the position of instruments in a room than either of the other 2 methods.

I once created a wish for this feature here: http://forum.fractalaudio.com/axe-fx-ii-wish-list/41513-true-stereo-reverb.html

As far as I know this cannot be achieved by using 2 reverb blocks (although paralell stereo can be achieved by doing this)... but its complicated so I might have been going about it the wrong way...

Now I just use true stereo bricasti IRs, which I prefer to the axe reverb anyway. http://www.samplicity.com/bricasti-m7-impulse-responses/

For realism, from the looks of it the ultimate stereo reverb would be Vienna MIR Pro. Holy crap that sounds cool. Check out the demo video! http://www.vsl.co.at/en/211/497/1687/2002/2010/1698.htm
 
Last edited:
as with all fx though... quite often it's more about it's application than the fx itself....
I have several reverbs [Lexicon, Sony Oxford, Space Designer, iZotope, TC Electronics, Waves, Native Instruments, plus the lower end reverbs in Logic, Digitech 2120 and a few others..]
some in that list are AU's, some are hardware, some are digital and some are convolution..
here's what I've found...
short ambient reverb [drums, riffing guitars etc] I really like Space Designer wood rooms
longer 'hall' reverbs for vocals is odd... for my voice I use the Lexicon, but for the singer in the DC Band [who I play for] the Oxford seems to work better..
DC's Zeta electric violin works really well with the Oxford..
my lead tones get on well with Space Designer and oddly enough Logic's Platinum Verb sounds great [although it's certainly not the best reverb out there.. it just works]..
my acoustics work really well with the Space Designer cathedrals and also the Oxford [although I'm going to explore Native Instruments more next time cos it's so beautiful and smooth]..
when I'm recording vocals I send to the 2120 and use the reverb on there [although I'm recording dry] and it's surprisingly very good...
if I didn't have some of the 'bigger hitters' in my reverb collection I'd be quite happy to use it in recordings..
for longer reverbs on the snare and toms you can get great results with Logic's Silver verb cos it's slightly trashy..
and the list goes on

moral of the story???? yet again it's all about application..
the nature of the voicing of the input signal.. how it reacts with the reverb.. the effect you're trying to achieve and how it sits in the mix...

I guess you could say the same about all fx....
which is why I have a ton of different compressors, eq AU's, delays etc etc etc......
in their own way they're all good.. even the ones you think are crap.. it's a matter of getting familiar with them and finding out what instruments / settings suit them best..

EDIT: maybe it's better to say that "one size don't fit all"
and that is the reason that I don't trust the reverb 'tests' above...
they are 100% wet and played via a very 'iffy' snare..
that very same reverb would sound seriously different with a riffing guitar, acoustic, vocals [as Nikki pointed out already], church organ, synth etc etc..
and then would sound different in the mix...

some of the example reverbs above would make you think "hell no"..
but in situ and blended with the right instrument may make you think 'fk yeah.."

Q - hands up all those in here that, when adding reverb to anything, create the reverb and fine tune the settings with just the 100% wet signal..??
and when mixing, how many of you sort out the reverb without the mix playing????

personally, my answers are hands down to both of those....
which means I have much less of an idea of what a 100% wet, soloed reverb sounds like..
as opposed to one that's in an applied situation and being fine tuned to the input signal and / or mix...
so personally, I don't really get much from those test recordings...
sorry guys.. not trolling here... just my personal point of view..
 
Last edited:
btw - I personally thought the Axe2 reverbs were rather nice when I tried them...
they seemed to tick all the right boxes...
and it's really hard to get a real sweet studio type reverb tone through a pair of 4x12 cabs...
I thought the Axe did really well in this respect..
 
Last edited:
the obvious way to go would be to have convolution reverbs in the axe, but this would obviously have implications as far as cpu usage goes...unless it's a new block which the user could choose to use if there was enough cpu headroom for it

i have to say that i tend to use the plex detune more than the reverb block for very long, dense reverbs, but with a few tweaks, the reverb block is capable of some really nice stuff. perhaps all it would take is for someone to create a new range of presets for it, which we could all save and use as fx blocks. i know it's somewhat limited, being mono in, but it might be worth a go, especially if they were based on some of the nicer hardware and software stuff out there. i think clarky might be in the best position here... :)
 
nikki-k said:
Reverb test?
"Holy Grail" test, IMO, is, and always will be, well recorded vocals.
If a 'Verb can stand up to scrutiny when used on a great vocal track... :D
Click to expand...

I think you are right.

I just wonder if the CPU power to pull that off is worth it in the AxeFx? A guitar / bass oriented box?

Buy yeah, good enough to use on the LEAD VOX, is the litmus test for reverbs. Hardware and plugs.

Richard
 
Radley said:
I would agree that the AFXII reverb is not totally up to current sonic standards, but I feel sure there will soon be an update, since reverb is 'the mother of all effects'...
Click to expand...

In my mind, I've never tried to compare the AxeFx verb to plugs or hardware units.

I've always compared the AxeFx verb to my older guitar rigs. In that dept., the AxeFx blows away the pedals I used to use.

But when I compare it to a good convolution plug, or the UA reverb plugs that I use in the box, yeah it's not there for sure.

Richard
 
the question is...does it need to be...

the axe can't be all things to all men and if i wanted a top notch reverb on a lead vocal, i wouldn't patch the axe fx in to do it...i'd go elsewhere and i'm sure most people would

that's not taking anything away from the axe fx...you just use the right tool for the job. the axe fx reverbs are plenty good enough to be used on guitars...live...which is what it was designed to do, i guess.
 
simeon said:
the question is...does it need to be...

the axe can't be all things to all men and if i wanted a top notch reverb on a lead vocal, i wouldn't patch the axe fx in to do it...i'd go elsewhere and i'm sure most people would

that's not taking anything away from the axe fx...you just use the right tool for the job. the axe fx reverbs are plenty good enough to be used on guitars...live...which is what it was designed to do, i guess.
Click to expand...

That's how I have looked at it and am completely happy with it for that purpose.

Richard
 
nikki-k said:
Reverb test?
"Holy Grail" test, IMO, is, and always will be, well recorded vocals.
If a 'Verb can stand up to scrutiny when used on a great vocal track... :D
Click to expand...

I wouldn't consider Vocals a holy grail test, just ONE of the many tests you could run on an algo, basically different types of source materials: Vocals, guitar, synthesizer, impulsive sounds (mallets, drums, etc). The particular test I posted is very pulsive - One is surprised at how many artifacts this can generate on some reverb algos. However, it's just one test and to me the Axe-Fx Passed.

It does show off the echo density which is what I was mostly interested in comparing, since to my ears Valhalla does a wonderful job at it.
 
simeon said:
the obvious way to go would be to have convolution reverbs in the axe, but this would obviously have implications as far as cpu usage goes...unless it's a new block which the user could choose to use if there was enough cpu headroom for it

i have to say that i tend to use the plex detune more than the reverb block for very long, dense reverbs, but with a few tweaks, the reverb block is capable of some really nice stuff. perhaps all it would take is for someone to create a new range of presets for it, which we could all save and use as fx blocks. i know it's somewhat limited, being mono in, but it might be worth a go, especially if they were based on some of the nicer hardware and software stuff out there. i think clarky might be in the best position here... :)
Click to expand...

I specifically avoided convolution reverbs as that imposes latency requirements that would be beyond what I would be willing to tolerate. That's what I'm comparing against other reverb simulators.
 
AlbertA said:
I wouldn't consider Vocals a holy grail test, just ONE of the many tests you could run on an algo, basically different types of source materials: Vocals, guitar, synthesizer, impulsive sounds (mallets, drums, etc). The particular test I posted is very pulsive - One is surprised at how many artifacts this can generate on some reverb algos. However, it's just one test and to me the Axe-Fx Passed.

It does show off the echo density which is what I was mostly interested in comparing, since to my ears Valhalla does a wonderful job at it.
Click to expand...

That is fair. Especially in an analytical context vs. a purely "mixing" context.
 
clarky said:
as with all fx though... quite often it's more about it's application than the fx itself....
I have several reverbs [Lexicon, Sony Oxford, Space Designer, iZotope, TC Electronics, Waves, Native Instruments, plus the lower end reverbs in Logic, Digitech 2120 and a few others..]
some in that list are AU's, some are hardware, some are digital and some are convolution..
here's what I've found...
short ambient reverb [drums, riffing guitars etc] I really like Space Designer wood rooms
longer 'hall' reverbs for vocals is odd... for my voice I use the Lexicon, but for the singer in the DC Band [who I play for] the Oxford seems to work better..
DC's Zeta electric violin works really well with the Oxford..
my lead tones get on well with Space Designer and oddly enough Logic's Platinum Verb sounds great [although it's certainly not the best reverb out there.. it just works]..
my acoustics work really well with the Space Designer cathedrals and also the Oxford [although I'm going to explore Native Instruments more next time cos it's so beautiful and smooth]..
when I'm recording vocals I send to the 2120 and use the reverb on there [although I'm recording dry] and it's surprisingly very good...
if I didn't have some of the 'bigger hitters' in my reverb collection I'd be quite happy to use it in recordings..
for longer reverbs on the snare and toms you can get great results with Logic's Silver verb cos it's slightly trashy..
and the list goes on

moral of the story???? yet again it's all about application..
the nature of the voicing of the input signal.. how it reacts with the reverb.. the effect you're trying to achieve and how it sits in the mix...

I guess you could say the same about all fx....
which is why I have a ton of different compressors, eq AU's, delays etc etc etc......
in their own way they're all good.. even the ones you think are crap.. it's a matter of getting familiar with them and finding out what instruments / settings suit them best..

EDIT: maybe it's better to say that "one size don't fit all"
and that is the reason that I don't trust the reverb 'tests' above...
they are 100% wet and played via a very 'iffy' snare..
that very same reverb would sound seriously different with a riffing guitar, acoustic, vocals [as Nikki pointed out already], church organ, synth etc etc..
and then would sound different in the mix...

some of the example reverbs above would make you think "hell no"..
but in situ and blended with the right instrument may make you think 'fk yeah.."

Q - hands up all those in here that, when adding reverb to anything, create the reverb and fine tune the settings with just the 100% wet signal..??
and when mixing, how many of you sort out the reverb without the mix playing????

personally, my answers are hands down to both of those....
which means I have much less of an idea of what a 100% wet, soloed reverb sounds like..
as opposed to one that's in an applied situation and being fine tuned to the input signal and / or mix...
so personally, I don't really get much from those test recordings...
sorry guys.. not trolling here... just my personal point of view..
Click to expand...

Well I think the point of the thread was kinda lost (i.e. this is not a thread about how to best apply the reverb block, Radley has very useful threads for that). Yes I agree this is just but one test as acknowledged above; but this is not how a thread about how a reverb sits in the mix.

When comparing reverbs, listeting to just their 100% wet output is very informative. Pulsive sounds bring out 2 things: the build up of echo density and how propense the reverb is to metallic resonances (this is mostly evident on algos that use allpass networks).
 
Last edited:
simeon said:
the question is...does it need to be...

the axe can't be all things to all men and if i wanted a top notch reverb on a lead vocal, i wouldn't patch the axe fx in to do it...i'd go elsewhere and i'm sure most people would

that's not taking anything away from the axe fx...you just use the right tool for the job. the axe fx reverbs are plenty good enough to be used on guitars...live...which is what it was designed to do, i guess.
Click to expand...

Sure I understand that, however for analogy purposes, one could have said the same about amp emulation a couple of years ago "The amp are good enough to be used on guitars...live". Yeah no doubt.
However imagine after that there was no push for improvement on the amp sim front...One could have just as easily said, "Use the right tool for the job, i.e. a real amp".
 
I really like the reverb in the Axe-FX II. It does everything I need it to, and sounds really natural to my ears at least.
 
simeon said:
the question is...does it need to be...

the axe can't be all things to all men and if i wanted a top notch reverb on a lead vocal, i wouldn't patch the axe fx in to do it...i'd go elsewhere and i'm sure most people would

that's not taking anything away from the axe fx...you just use the right tool for the job. the axe fx reverbs are plenty good enough to be used on guitars...live...which is what it was designed to do, i guess.
Click to expand...

There is no harm in asking.

I also see no problem in describing what the Axe II is in these comparisons. I'm sort of in with those who are in the 'practical application' of 'verbs. I'm sure there are better ones, but the Axe 'verbs are better than 90% of units out there, so they work great for me. I'm not sure that they were designed to compete with units or plug-ins whose sole purpose is 'verb.

Either way, I watch with interest.

Ron
 
You must log in or register to reply here.
Back
Top Bottom