Jay's 1024 cabs?

[AlbertA]

All I gotta say, it's the 1024 points IR, really do make a difference for me in the low end quality. I don't know if it's placebo, but immediately notice the enhanced definition (I don't know how else to describe it) using same patch only difference was 512 point vs 1024 point cabs. Just flipping back and forth between them.

 

[scarr]

I get your point that the discontinuity is the big concern here. In a theoretical world, if that were somehow eliminated,

The first half of a 1024-point IR ends on an unknown value. The second half begins at a nearby value such that the combined function is continuous. Eliminating any possible discontinuity would require forcing the first half to end at a known value. The only defensible choice for that value is zero, and the only technique for forcing the function to zero at the end of the IR while minimizing undesirable artifacts is data windowing. As you've acknowledged, concatenating a spurious second half with a first half that has been windowed is not a productive exercise.

Implementing a statistical tail extension would potentially require more resources than just extending the IR length to 1024 points, and we have a pretty good idea that that isn't feasible.

Good point. I guess it's just one of those things there is no getting around.

Thank you for taking the time to go over all this with me. I really appreciate it!

 

[Mike Snider]

Thanks for having this discussion. Fascinating.

 

[m lebofsky]

I can only hope Cliff is watching and understands that 1024 stock Axe IRs in version 6 would make a lot of people feel real good ....

 

[Jay Mitchell]

I can only hope Cliff is watching and understands that 1024 stock Axe IRs in version 6 would make a lot of people feel real good ....

He's known that for some time and has spent a nontrivial amount of his time pursuing that. The problem is that, unless some real economies of resources can be realized elsewhere, there isn't enough capacity (cpu and/or RAM, I don't know which) to implement them.

 

[m lebofsky]

[quote="m lebofsky":fmrnwcg4]I can only hope Cliff is watching and understands that 1024 stock Axe IRs in version 6 would make a lot of people feel real good ....

He's known that for some time and has spent a nontrivial amount of his time pursuing that. The problem is that, unless some real economies of resources can be realized elsewhere, there isn't enough capacity (cpu and/or RAM, I don't know which) to implement them.[/quote:fmrnwcg4]

Personally, I would give up other resource intensive options (i.e., multiple amps per patch) if it meant better IRs.

 

[Jay Mitchell]

Personally, I would give up other resource intensive options (i.e., multiple amps per patch) if it meant better IRs.

Well, there's a perfect example: I wouldn't consider sacrificing Amp 2. I make use of it in several presets, and I haven't found a way to get close to the same result with just one amp block.

 

[AlbertA]

I think Cliff has mentioned two limitations from his posts:

The Standard Axe-Fx doesn't have enough RAM to hold another new type of block (this is my speculation based on his recurring comments, also remember all blocks are instantiated at all times, which means all RAM resources will be consumed, or at least that's how I currently understand the architecture). Therefore it's probably difficult to add a "new 1024 point block". OR it could be that the RAM resources required to give the existing cab blocks a 1024 option exceed what's currently available.

I think the above limitation may not apply given the comment below:
http://www.setbb.com/axefx/viewtopic.ph ... xefx#52308
He mentioned a reworked cab block that allowed 1024 IR's but added an additional 1ms of latency which he hated (I personally could live with it).

 

[m lebofsky]

[quote="m lebofsky":2sllixrq]Personally, I would give up other resource intensive options (i.e., multiple amps per patch) if it meant better IRs.

Well, there's a perfect example: I wouldn't consider sacrificing Amp 2. I make use of it in several presets, and I haven't found a way to get close to the same result with just one amp block.[/quote:2sllixrq]

Just an example. I am sure different people feel strongly about different features/capabilities. 1024 IRs could be used to help all of my patches. The second amp is used to help perhaps 2 of my current patches. I have been putting off changing all of my patches to include the current implementation of split 1024s because I have been hoping for 6.0 to address the issue.

I do remember him mentioning the development of 1024 IR's that added 1ms of latency. I would take the added 1ms. YMMV.

 

[Jay Mitchell]

Just an example. I am sure different people feel strongly about different features/capabilities.

Which is why taking anything away once you've given it is off the table. That would be a major strategic blunder. You never know how many users will suffer from any change that removes a feature.

1024 IRs could be used to help all of my patches.

They certainly help all of mine.

I have been putting off changing all of my patches to include the current implementation of split 1024s because I have been hoping for 6.0 to address the issue.

I certainly don't have any special information, but everything I've seen on this subject so far says "don't hold your breath." If you set your Axe-Fx up for split IRs, it will be a simple matter to make use of 1024s if they ever appear.

 

[Dutch]

I certainly don't have any special information, but everything I've seen on this subject so far says "don't hold your breath."

Bummer.
You'd think if one could fabricate faux-1024 cabs as you have been doing here, using multiple cab blocks and delays, the Axe-FX would be capable of real 1024 cabs and that that would even be more economical than aforementioned construction. Does this construction actually add to the latency?

I was so hoping for some improvement in the cabs.

My unsubstantiated gut-feeling is that it's the cabs that are the one thing in the AxeFX that could use some improvement. That or my monitors. And my playing of course...

Well, maybe there will be improvement in the cabs some other way. Like I could get lessons or something...

 

[m lebofsky]

[quote="m lebofsky":l63e89gc]Just an example. I am sure different people feel strongly about different features/capabilities.

[/quote:l63e89gc]Which is why taking anything away once you've given it is off the table. That would be a major strategic blunder. You never know how many users will suffer from any change that removes a feature.

m lebofsky says: Perhaps taking away things could be an option, thus allowing those who want the 1024s to use them. Many different ways to skin a cat so to speak. Just trying to make it easy/convenient for people (like myself) to use 1024s.

 

[m lebofsky]

I certainly don't have any special information, but everything I've seen on this subject so far says "don't hold your breath."

Bummer.
You'd think if one could fabricate faux-1024 cabs as you have been doing here, using multiple cab blocks and delays, the Axe-FX would be capable of real 1024 cabs and that that would even be more economical than aforementioned construction. Does this construction actually add to the latency?

I was so hoping for some improvement in the cabs.

My unsubstantiated gut-feeling is that it's the cabs that are the one thing in the AxeFX that could use some improvement. That or my monitors. And my playing of course...

Well, maybe there will be improvement in the cabs some other way. Like I could get lessons or something...

I spent the past 2 days comparing my 2 QSCs to my 2 Bogner 2x12 cabs. Its the cab IRs (based on my ears).

 

[Mike Snider]

m lebofsky says: Perhaps taking away things could be an option, thus allowing those who want the 1024s to use them. Many different ways to skin a cat so to speak. Just trying to make it easy/convenient for people (like myself) to use 1024s.

I think that has been discussed at length on the old forum and the resolution is that would not happen with the current standard or Ultra. All good ideas....but evertime it's discussed it comes back to the same resolution. Cliff may suprise us as usual.

In the meantime, enjoy the 1024's which are available...and enjoy tweaking.
I LOVE MY AXEFX!

 

[Amphibian]

Seriously guys, once you've tried the 1024 point IR's there's just no turning back. Since I tried the 1024's I have basically been playing this one single preset with the Dumble and the new cab Jay made. I actually think it's twice as good and convincing! I would personally compromize a whole lot to have the 1024 as stock configuration, latency is however not an option, IMHO! My hope is that Cliff will have another devine epiphany before 6.0! I don't know if it's at all possible though...

Cliff, isn't it possible to hide the algorithm for the 1024 point method in a single mono cab block? Just take the concept of the 10 ms delay and panning and add it to a new block? IMHO, I think it's strange that you can't find a way to make it work when it's working so good with the method invented by using the delay block!?

 

[scarr]

Cliff, isn't it possible to hide the algorithm for the 1024 point method in a single mono cab block? Just take the concept of the 10 ms delay and panning and add it to a new block? IMHO, I think it's strange that you can't find a way to make it work when it's working so good with the method invented by using the delay block!?

Giving the Cab a switch for 2x512/1x1024 is on the Wish List, although I don't know if there's a thread for it yet.