HOTONE AMPERO STOMP / STAGE NOW " IMPORT " NAM FILES

[Mike Oliphant]

WOW now this is HUGE! I hope the NAM files use way less CPU than their clone files.

This using the NAM file as a basis for creating one of their clone captures, so you would expect that it would use the same amount of CPU.

 

[JCW308]

This using the NAM file as a basis for creating one of their clone captures, so you would expect that it would use the same amount of CPU.

Ahhhh true.

 

[BenIfin]

Here's a comparison.

Signal chain:

Guitar (Carvin C66 with Suhr Doug Aldrich pickups) -> Hotone Ampero 2 Stomp -> USB -> Macbook Pro 16" M2 Max -> Logic Pro.

Capture:

BluGuitar Amp 1 Mercury Edition Classic channel -> Bluetone Loadbox -> Audient EVO 8 audio interface.
NAM 1000 epoch capture made using Tonezone3000.

Recording setup:

• Input 7 (Ampero 2 DI signal) -> NAM plugin (mono -> stereo) -> Bus with ML Sound Lab MIKKO 2 cab sim.
• Input 1/2 (Ampero 2 main outs) -> Same bus with cab sim.

I monitored the DAW output from the A2 Stomp headphone out

With a bit of reverb from UA Capitol Chambers, and UA LA-2 output compression:


Same clips, bone dry:


Thoughts:

The low end is totally different on the Ampero 2. I need to try other captures, maybe 100 epoch captures or something. The straight NAM capture in the plugin sounds much clearer.

Hey Laxu

Thanks for posting.

Your clips are really interesting because if you listen "through" the low end hump, the mids and the highs sound virtually identical.

Perhaps there is a quirk (?) in how you did the comparison ? I don't know ?

I only say this as Hotone would clearly have tested this before release and I cant imagine they would have "stuck" with an import-and-convert process that "added" so much low end to an existing NAM file ?

Even the Sonuloab NAM Player that Leo Gibson Demo'd, which also does its own conversion, sounded almost identical to the NAM Player, but fell a bit short in the null test - which of course is neither here nor there

Thanks for trying and persisting !

 

[laxu]

I'm not entirely sure how you set it up, but if I read correctly you used the interface of the Ampero for all of this, so instead you could maybe try and compare by recording a di track through your normal interface into your DAW and then reamping that through both the Ampero and the NAM plugin and seeing if that helps. Also, I would leave out any post processing as that's not really helpful to compare (eg. reverbs etc).

The way I see it, my setup should be as close to like for like as the Ampero acts as the hardware. Running DI from my audio interface to the A2 Stomp would add extra conversion etc.

To me the NAM capture in the DAW performs about right, even though I haven't measured input levels for the A2.

I will try again when I have the time.

I posted the dry tracks for comparison as well.

 

[Dom Mcsweeney]

The way I see it, my setup should be as close to like for like as the Ampero acts as the hardware. Running DI from my audio interface to the A2 Stomp would add extra conversion etc.

To me the NAM capture in the DAW performs about right, even though I haven't measured input levels for the A2.

I will try again when I have the time.

I posted the dry tracks for comparison as well.

Yeah no biggy, I don't think one way is better than the other. Simply thinking out loud for another way to test.

 

[DLC86]

Here's a comparison.

Signal chain:

Guitar (Carvin C66 with Suhr Doug Aldrich pickups) -> Hotone Ampero 2 Stomp -> USB -> Macbook Pro 16" M2 Max -> Logic Pro.

Capture:

BluGuitar Amp 1 Mercury Edition Classic channel -> Bluetone Loadbox -> Audient EVO 8 audio interface.
NAM 1000 epoch capture made using Tonezone3000.

Recording setup:

• Input 7 (Ampero 2 DI signal) -> NAM plugin (mono -> stereo) -> Bus with ML Sound Lab MIKKO 2 cab sim.
• Input 1/2 (Ampero 2 main outs) -> Same bus with cab sim.

I monitored the DAW output from the A2 Stomp headphone out

With a bit of reverb from UA Capitol Chambers, and UA LA-2 output compression:


Same clips, bone dry:


Thoughts:

The low end is totally different on the Ampero 2. I need to try other captures, maybe 100 epoch captures or something. The straight NAM capture in the plugin sounds much clearer.

Definitely different sounding. IIRC the Ampero runs at 44.1 kHz right? As someone mentioned about the sonicake, this different tone might be the consequence of the conversion not taking into account the different sample rate of the original model.
One test I would do to verify this would be to load a 44.1 kHz NAM model (it can be done quite easily with some trickery) and see if it sounds the same in that case.

 

[DLC86]

And in the meantime also Valeton adds NAM support to all the GP200 line

https://m.facebook.com/story.php?story_fbid=pfbid0atSv5ftwaBurGt5ZkpCowSwk2qGorqXGk3mnzJGh9jocQG1Qex8fHn3ZX2pqB4mil&id=100083540640039

 

[Dom Mcsweeney]

They're all owned by the same company (Valeton, Hotone, Sonic Cake)

And, from what I have read so far (and heard) none of these units support 'nam files' at all, instead they do a tonematch or something similar. A lower quality capture of a capture.

If I was heavily invested in making free/commercial nam packs, I'd be pretty pissed off if those were one day cloned and uploaded to Hotone's cloud server.

I think also, this really doesn't help anyone in the industry to move forward - it just undermines any company doing genuine work in this area (such as Dimehead, or anyone else who have added NAM support in their software/hardware).

 

[BenIfin]

And in the meantime also Valeton adds NAM support to all the GP200 line

https://m.facebook.com/story.php?story_fbid=pfbid0atSv5ftwaBurGt5ZkpCowSwk2qGorqXGk3mnzJGh9jocQG1Qex8fHn3ZX2pqB4mil&id=100083540640039

Gee's ...

Sooner or later ... my money is on sooner ... one of these 2nd Tier makers is going to get their "import and conversion" of NAM files to be "transparent" - as in sonically and audibly indistinguishable to the NAM Player

Interestingly (?) the Valeton FW notes refer to "importing" NAM Captures .. no reference to any "conversion" or "proprietary stuff being done" (?)

Accidental omission (?) Deliberate omission (?) Truthful and accurate (?)

 

[BenIfin]

They're all owned by the same company (Valeton, Hotone, Sonic Cake) And, from what I have read so far (and heard) none of these units support 'nam files' at all, instead they do a tonematch or something similar. A lower quality capture of a capture.

^^ Not saying you are wrong or right, but your Googling / research skills are way better than mine as I cant find anything anywhere either way , supporting or disputing this ??? or anything at all about how/what "conversion processes" they may or may not use ???

 

[Dom Mcsweeney]

^^ Not saying you are wrong or right, but your Googling / research skills are way better than mine as I cant find anything anywhere either way , supporting or disputing this ??? or anything at all about how/what "conversion processes" they may or may not use ???

I could be wrong in the macro details of this, but they aren't hosting NAM files - i'm quite certain.
The Hotone forum states that it's converted, and, after that they say they send an impulse through the targets being cloned:
(from here: https://community.hotoneaudio.com/forum/discussionDetail?id=1893)

Someone in the NAM fb group commented this under the post about Hotone:

----

And when I say 'heard' I simply mean listening to the previous sound bites earlier in this forum thread by Iaxu.

------

All of my thoughts are simply opinions, not fact, but I remain highly doubtful of any company going in this direction and wish they would instead just do it right. Stating things like "we support NAM models" - when really they don't - isn't fair on any consumer, or any competitor in the industry.

 

[Dom Mcsweeney]

^^ your Googling / research skills are way better than mine

I try to never use google, I'm not really into being their product and I don't like all the advert based results.
I mostly use Duck duck go, but, for any and all of this - I just look in the places I expect information.
Eg. their forum, their socials, the NAM fb group / comment sections / here / other places that are also discussing things.

I don't claim to be 100% right about anything, and I'd welcome being corrected on any of what I've said. =)

Also, it's so vague and unclear, I don't think anyone will get the full information and I doubt the company will explain it transparently any time soon.

Someone like Leo Gibson will hopefully do us all a service and make a deep dive when he has time.

 

[[Nathan]]

If something is using a process to “photocopy” or “downscale” a capture file then it’s an instant pass for me. If we knew more about the conversion process I could maybe open my mind up to the idea but we can take some uneducated guesses that it’s just using some approximation method and “hey if sound comes out then great”!

I would be just as unthrilled if this did the same thing to tonex, Kemper or qc captures. A photocopy of a photocopy isn’t interesting, might as well just use modelling and get better results. I’m guessing this is their way of getting a cool spec as a sales point. Hopefully this just leads to proper native NAM support for them in the future, that would be interesting.

FWIW the ampero stuff looks good on its own, I’d use the rest of the unit and just skip this implementation of profiling. The units seem pretty good without this stuff.

 

[2dor]

If something is using a process to “photocopy” or “downscale” a capture file then it’s an instant pass for me. If we knew more about the conversion process I could maybe open my mind up to the idea but we can take some uneducated guesses that it’s just using some approximation method and “hey if sound comes out then great”!

I would be just as unthrilled if this did the same thing to tonex, Kemper or qc captures. A photocopy of a photocopy isn’t interesting, might as well just use modelling and get better results. I’m guessing this is their way of getting a cool spec as a sales point. Hopefully this just leads to proper native NAM support for them in the future, that would be interesting.

FWIW the ampero stuff looks good on its own, I’d use the rest of the unit and just skip this implementation of profiling. The units seem pretty good without this stuff.

They're probably doing some weight quantization or pruning if the unit's doing inference.

 

[Jkist7]

They're all owned by the same company (Valeton, Hotone, Sonic Cake)

And, from what I have read so far (and heard) none of these units support 'nam files' at all, instead they do a tonematch or something similar. A lower quality capture of a capture.

If I was heavily invested in making free/commercial nam packs, I'd be pretty pissed off if those were one day cloned and uploaded to Hotone's cloud server.

I think also, this really doesn't help anyone in the industry to move forward - it just undermines any company doing genuine work in this area (such as Dimehead, or anyone else who have added NAM support in their software/hardware).

If you are talking about this affecting profile vendors, I don't sympathize. The idea of creating captures of an amp is already kind of a moral grey area. You can rent a $10,000 amp (or buy from Amazon), make a bunch of captures of it, return it, and then turn around and sell those captures on your website indefinitely and make profit that requires very little upkeep.

There are a ton of vendors out there doing exactly this. Does that seem fair to the original amp builder to you?

Back on topic, yeah the Hotone converted NAM profiles seem more scooped and hyped in the lows. Its not mega, mega far off, but hopefully they can improve the process.

 

[Dom Mcsweeney]

I think there is definitely merit when people do buy amps, or rent them. That is all positive.
Making captures of amps will increase the brand awareness - I know alot of people (including myself) that have either bought an amp after trying a NAM model/or similar/ or have simply bought an amp so they can make more captures.
Besides, many amp companies do this themselves by releasing their own myriad digital products.

 

[Jkist7]

I think there is definitely merit when people do buy amps, or rent them. That is all positive.
Making captures of amps will increase the brand awareness - I know alot of people (including myself) that have either bought an amp after trying a NAM model/or similar/ or have simply bought an amp so they can make more captures.

I think there are far more out there that figure the captures are close enough, and decide against buying an entire amp. But either way, neither of us have hard data.

Besides, many amp companies do this themselves by releasing their own myriad digital products.

Precisely. If I was an amp builder, I would feel that I HAVE to put my own digital versions out there, that way I'M the one making money off the thing I originally created anyways. If I don't, someone else will, and then none of that money is coming to me. They don't have to pay royalties or anything of the sort.

 

[Dom Mcsweeney]

Yeah, I dunno - the industry has been shifting over to digital modelling instead of big loud amps for many many years, way before NAM/Kemper/Tonex/NDSP etc.

Digital modellers of all kinds have advanced, and they all do something along the lines of 'Martian800' etc.
I'd argue that black box modelling is more supportive of amp builders than 'white box', but I think the amp industry is still fairly healthy.
But yeah, who knows.

 

[Dom Mcsweeney]

...and if people with a Hotone/Valeton or whatever just take what has already been captured by NAM and then they clone onto their Hotone, then share those models on other cloud servers... they definitely aren't supporting the people who make amps, nor are they supporting people who have made packs, etc.

 

[Jkist7]

Yeah, I dunno - the industry has been shifting over to digital modelling instead of big loud amps for many many years, way before NAM/Kemper/Tonex/NDSP etc.

Digital modellers of all kinds have advanced, and they all do something along the lines of 'Martian800' etc.
I'd argue that black box modelling is more supportive of amp builders than 'white box', but I think the amp industry is still fairly healthy.
But yeah, who knows.

I think the arguement is a little different with "amp models" versus captures. As a user, you don't necessarily assume the "Martian800" is a 1 to 1 clone of the JCM 800, but more of their interpretation of that sound or amp.

A capture, by its definition, is supposed to be a 1 to 1 recreation of an amp at given settings.

But yes, it all gets very messy. I'm just saying that its ironic to me to point the finger of judgement at Hotone, then hold profile vendors on some higher pedestal.