SaschaFranck

What's the specs of your machine? And would you care to check how long it takes to load a test project? And fwiw, why are you still on Logic 10.4.7?

I am very obviously using hardware monitoring all the time (basically, when you switch the Helix on, you're monitoring through hardware) and I never expected the Helix to be as suitable for software monitoring as a dedicated audio interface. Yet, there's always the moment when software monitoring comes in handy as an additional option. For example, checking IRs is cumbsersome on the Helix itself (due to the rather small number of IR slots) but works really nicely inside a DAW. I may as well want to try out some other amp sims just to have a brief look at what things would sound like through them. All this is no fun at all at 17ms latency - and it's even so much you can't clearly judge about things such as IRs anymore because the playing feels so distracted. In addition, 10ms son the output side alone are good enough to make you compensate too much by playing early when using software instruments (and with those, hardware monitoring is no option). Whatever, I'm not asking for RME league latency figures or whatever - but I'm asking for something a little better than what has to be considered almost the worst on the market. @Digital_Igloo, are you guys ever looking into that?

Why would an authorization problem cause slow loading times? And fwiw, I have seen it act pretty slow on other Macs already as well.

Made a stress test some days ago - I can load well over 30 instances of HXN (comp, drive, amp, cab/IR, EQ, some FX in each instance) easily, but project loading time is horrible. With Logic already opened in the background, a 30 instances project takes over 2 minutes to load the project. More than even the most demanding projects using large sample based patches - and it's not that HXN would have anything to load but its UI and a few kb-sized IRs.

Yeah, thanks. Same observations. Especially under macOS, the Helix driver is horribly bad. IMO it should be at least good enough to have an occasional testrun of, say, a software amp sim plugin. Something such as 8-10ms at 44.1 and 64 samples would likely be sufficient. At almost 17ms it's a pain to deal with when trying things like that. As said in that thread, the figures we're getting right now are those from 10+ year old cheap interfaces. Even my almost 20y old Emagic EMI 6/2 is doing way better.

I do have a pretty decent Zoom UAC-2. But using the Helix is *way* more comfortable as I can grab the recorded signal from anywhere in the signal chain, don't need to go through additional ADDA conversion and what not. And seriously, I really don't expect "great" drivers - but they could at least be as good as what I'm getting through Core Audio. It's like that with pretty much any interface I ever tested, Core Audio usually is quite sufficient but dedicated drivers almost always are somewhat better. With the Helix it's the complete opposite. It's also worse than what you get from Fractal, Boss and - well - even cheap Zoom, Mooer and Hotone units. In a nutshell, we're talking about the worst drivers in any guitar modeler/MFX unit. If I was Line 6, I wouldn't want that.

Check. I learned to never trust reported driver numbers. At least for a while (admittedly some time ago) they were often plain wrong, especially for lower-end-ish units. It's still among the first things I do whenever I'm really interested in an interfaces performance.

As said, I am not expecting "professional audio interface" results out of the interface portion of the Helix. But I'm expecting a little more than the worst possible - which is precisely what we have right now. Again as said, even things such as a uber-cheap Behringer U-Phoria interface beats the Helix by several miles when it comes to low latency. And fwiw, you'll have a tough time finding any interface doing as bad as the Helix on the GS thread (which I am following ever since it started), even the oldest and cheapest ones included. Oh, and yet again as said: Even in class compliant mode, performance is *way* better. Which speaks volumes about the driver development.

Ok, I'm wondering now. Or venting off. Or whatever... Latency of the Helix as an audio interface is all the way on par with the worst of the worst interfaces on the market. Under macOS, at 44.1kHz and 64 samples buffersize, roundtrip latency is 16.9ms (measured manually - but also reported by Logic). Seriously, that's as bad as it gets. And here comes the kicker: When using it without the dedicated driver as a class compliant interface through Core Audio (which is unfortunately causing some issues here and there and not a parallel option anymore for whatever weird reasons since firmware 2.8 I think...), the resulting latency figures are *way* lower, IIRC (can't be bothered to delete the driver at the moment) something around 9ms. Usually, whenever audio interfaces come with their own driver, the RTL figures are better compared to using Core Audio (or ASIO4All under Windows), with the Helix it's the opposite. IMO this shows a clear lack of effort in the driver development department. Even my Zoom G3 comes up with less latency, the Boss GT-10 too. Add to this that, while causing some issues, things such as crackles or so in recordings aren't happening when using the Helix in class compliant mode (these would usually be a reason to raise the internal "safety buffers"). I'm absolutely fine with mediocre driver performance on a unit which isn't meant to be a professional audio interface in the first place - but as being an audio interface is at least a rather large part of the "Helix idea" (otherwise we wouldn't see all the USB routing options and such), it should defenitely not deliver the worst performance, either. To give you an idea: Most interfaces in the 100 bucks realm offer around half the roundtrip latency at the same settings, my €200 Zoom UAC-2 goes down to 5.9ms. As is, it's completely impossible for me to have any fun with, say, a software amp sim. Heck, it's so bad that discriminating keyboard players would complain about just the output latency (around 9ms plus input MIDI latency), let alone drummers triggering some samples. Time to give the driver development some love I'd say.

Hm, ok, didn't know this was about arrangements and such as well (I thought it was a plain sound issue). With two overplaying keyboarders, things are really getting tough. Anyway, in that case, I highly recommend to just listen to the keyboards and try to find your space. Basically, this can happen tonally, frequency-wise and rhythmically. In each of these areas, you have several options at your disposal. Keyboarders playing huge chords? Fine, go for way smaller chord structures. Triads work pretty well quite often, especially as you can play inversions all over the neck with ease. But maybe you'll even have to reduces yourself to double stops or single notes. Again, these can be played all over the neck, so in case the keyboards are covering the lower registers, go way up to, say, 10th position and above. Are you allowed to rock out using, say, powerchords a little? Fine too, because this is where you can really set yourself apart from keyboarders. Using a marshall-esque type of amp will help you to cut through. Are the keyboards covering the low and high end? Try to find your frequency spot in the midrange area. What's going on rhythmically? This always requires special attention. Apart from fat riffs and directed/organized spots (such as accents and stops) there's zero need to double a rhythm played by a keyboarder. Rather go for something complementary. In case they're playing quarternote based stuff, go for 8th notes inbetween. In case they're playing busy stuff, play less (and vice versa). In case they're playing long sustained chords, go for something more accented. Etc. Oh, and talk to the guys. "Are you going to play that line? Otherwise I could". Stuff like that. And fwiw, as has been said already, even if pretty much each and every P&W player wants to sell you their "washed delay/verb" patches 'n tricks, especially with two keyborders you might be better off using less effects. Keep those for the moments when you're exposed. In a full mix, they'll either get lost or help to create an undefined mess.

But the Global EQ is adjusting all patches at once. So things can done at a soundcheck. Impossible with individual EQs in case you're using more than just 2-3 patches. You can still transfer the Global EQ settings to a patch EQ later on.

I'd just set up a medium broad mid band in the global EQ (Q around 1) with a moderate boost (something like 3dB) and experiment with the center frequency.

You don't want one amp to run into another. What you need to do is to cable things up parallely.

Hm, wikipedia still states something like 32ms of latency when using aptX via BT. Which is great for BT but still completely useless for realtime playing.

Uhm. I haven't seen (or heard, or played through) a single BT device that would not introduce horrible latency for any kind of realtime tasks. Which headphones are you using?

A) Nobody wants to replace the stable connection options of the Helix. B) Both Wifi and Bluetooth are used in a lot of professional products. And fwiw, there's quite some reports about problematic USB ports (the little plastic thingy inside being broken). So much about a stable connection.

They could (and IMO should) have added some wireless control options to the entire Helix family. Along with some (possibly feature reduced) editor for mobile platforms. Bluetooth would likely be sufficient for any control duties. And yes, it's been well established in 2015 already, hence I think it's a massive oversight. And one of the reasons I almost went with a GT-1000.

Thanks - but it hasn't gotten a dedicated MIDI in, so it must happen through the mini USB port. Which device are you using to control it through that?

I was meant to post this before but always forgot. Anyway, here goes: Out of the plugins I'm using (which are quite a plenty), HXN is by far (!) the slowest to load, regardless of whether I force it to come up with an empty patch (via Logics library presets) or a completely stuffed one. Heck, even some rather large Kontakt instances come up faster - including sample loading! Seriously, there's no other plugin even remotely as slow. I don't think there's anything to do about it, but I think this could be massively improved - it's not that HXN would be all that much more complexed than others. Fwiw, I'm using a 5,1 2x2.66 Mac Pro. And yes, I'm aware it's a little dated, but that's rather irrelevant because, as said, other plugins aren't even close in terms of loading time.

Sorry, missed the Blackhole part - yeah, looks every bit as appropriate.

Really, Soundflower will do exactly what is asked for.

No, aggregate devices won't help as they don't show up as possible inputs in Quicktime (just the physical ins, which is irrelevant for the playbacks).

You might want to check out Soundflower, which is a virtual audio cable kind of thing (fwiw, it's free). You can send pretty much anything into it and then use its output as an input for Quicktime.

The weird thing with all this being that he's mentioning Audacity as well - from all I know, Audacity is not recording MIDI data at all (at least my pretty much up to date version isn't).

You must be doing something wrong here because at least in Garageband, you can't record MIDI onto an audio track. Make sure you create a new track and select one of the audio track options.