FM9 Firmware Version 5.01

I've been holding on to it since december of last year. I am really hoping it works out. I realize that it won't likely be as gapless as the big brother, with the CPU difference being what it is. As long as it is significant, I'm in. In a perfect world I would keep both. My original plan was to use the rack for the bigger/important gigs, and the fm9 for the smaller stage/short set/rehearsal scenarios. The CPU you get back from using cablab 4 ir multiple mic blends more than makes up for what the gapless eats up.
Just to be clear, it's an audio gap, not latency...
Sorry, maybe my English is not so good to understand the difference. For me, the term 'audio gap' describes a more significant issue than 'latency,' and I don't think it's a very big problem. The 'gap' simply takes too much time.

For me (I am a computer engineer, too), the Audio Gap when switching scenes seems to be more of a design issue rather than a result of CPU differences. I genuinely believe Fractal should produce a product with the same CPU power as the Axe-Fx III but still be portable for live performances. The CPU limitations do restrict some creative possibilities for me (which were not easy to achieve on the Boss GT-100 and required various workarounds). I can make them easily achievable on the FM9, but only being hindered by some CPU limitations.
 
Last edited:
I've been holding on to it since december of last year. I am really hoping it works out. I realize that it won't likely be as gapless as the big brother, with the CPU difference being what it is. As long as it is significant, I'm in. In a perfect world I would keep both. My original plan was to use the rack for the bigger/important gigs, and the fm9 for the smaller stage/short set/rehearsal scenarios. The CPU you get back from using cablab 4 ir multiple mic blends more than makes up for what the gapless eats up.
How much extra CPU is gapless taking up on the Af3mkiiT?
 
I think they say around 1.5%ish. I have dropped my IR usage at least 10-15% with the blends, however. Instead of running 4 dyna cab slots, I now have 2 cablab 4 blended IRs. They can sound pretty much the same, but I think better.
 
  • Like
Reactions: rog
I posted in Axe III thread about the gap difference between the III and FM9. Previously I was able to work around it but now that the III is essentially gap less my FM9 feels limited in a sense. In my band I am jumping around scenes quite a bit. It frees me from having to anticipate the gap. Things flow easier on stage. I am sure Cliff will get it over to the FM9. He is the wizard. :)
 
I have played several gigs with my Fm9T since I got it around Dec of last year. It sounded great and did the job well. I must say, however, that the difference in audio gap between the Fm9T and the Af3mkiiT is just too much. I may end up selling the fm9T to get another Af3mkiiT. I love the convenience of it, but it is a drag. I am really holding out to see if gapless makes it over before I decide. For now it is a great rehearsal tool and backup just in case something happens. It has not gotten much use, I hate to say.
Chances are gapless will come sooner rather than later. With 2 amp blocks, have you tried working around it?
 
For someone who has been performing live with a Boss GT-100 for over 10 years, the scene switching latency of the FM9 is quite challenging to accept. I really need gapless switching.
Of course, I'm currently using a clean tone amp with a single drive block and an high-gain amp for gapless switching. However, what I really desire is to use different amps to achieve a low-gain effect, and even high-gain amps.
It seems like I might need three different amps to meet my requirements if I can't change scene in gapless....
Nothing wrong with the BOSS. If you feel the models in it do it for you and it serves your needs, by all means use it. Many of us are here because we feel the modeling is unmatched, along with build quality and customer service. That matters, too.
 
Hi, last week I received my Fm9 turbo and I noticed a bit of latency. It was odd because having owned the Axe FX standard many years ago and the Kemper more recently, I never cared about the latency those units had, so I assumed Fm9 would be on par if not better. I tested it and I measured a 4.6 latency on the Fm9. The amp block is adding almost 3ms of latency. The guys at G66 tested their units with betters tools and confirmed my findings, they also said that this must be a firmware thing and it will be fixed soon.

Thank you!
 
Hi, last week I received my Fm9 turbo and I noticed a bit of latency. It was odd because having owned the Axe FX standard many years ago and the Kemper more recently, I never cared about the latency those units had, so I assumed Fm9 would be on par if not better. I tested it and I measured a 4.6 latency on the Fm9. The amp block is adding almost 3ms of latency. The guys at G66 tested their units with betters tools and confirmed my findings, they also said that this must be a firmware thing and it will be fixed soon.

Thank you!
I might ask how far away from your cabinet you stand? Every 12" introduces just under a millisecond delay from the time the sound wave is produced by the speaker to when it hits your ear. 4.6ms of latency is about the same as standing 4-1/2 feet away from your cabinet. I'm not sure there's an issue here, and I certainly don't notice any latency while playing.
 
Hi, last week I received my Fm9 turbo and I noticed a bit of latency. It was odd because having owned the Axe FX standard many years ago and the Kemper more recently, I never cared about the latency those units had, so I assumed Fm9 would be on par if not better. I tested it and I measured a 4.6 latency on the Fm9. The amp block is adding almost 3ms of latency. The guys at G66 tested their units with betters tools and confirmed my findings, they also said that this must be a firmware thing and it will be fixed soon.

Thank you!
How did you test it?
 
I might ask how far away from your cabinet you stand? Every 12" introduces just under a millisecond delay from the time the sound wave is produced by the speaker to when it hits your ear. 4.6ms of latency is about the same as standing 4-1/2 feet away from your cabinet. I'm not sure there's an issue here, and I certainly don't notice any latency while playing.
You're absolutely right. And players like Guthrie Govan, Steve Stevens, Alex Lifeson and more don't seem to mind.

The only problem is when latency adds up.
I've played tube amps for 20 years and often on stages where the speaker was up to 6 metres behind me. That was a strange feeling and so wedges had to help out here. In the All Analogue era, it was just a question of sound. With the first digital consoles, wedges also had phase problems.

But what would have happened if I had used modellers instead of analogue amps back then? These 4 - 8 ms would have been added to the natural delay time and would have worsened the playing feel. So the comparison with the distance to the cabinet is technically correct, but is practically useless.

Nowadays I mostly play in-ear. And that is the right way. The days of 4x12 speakers and the destruction of your own hearing are over.
And I initially had an interesting problem with in-ears. The sound was too direct for me. When you've been used to natural latency all these years, the transmission in "zero time" was strange. In this respect, I don't mind the latency - as long as it doesn't get too high.

Ultimately, it's a question of application. I also have three different guitar transmitters in use. A Sennheiser ew100 G4, a Sennheiser ew-D and two Line6 G70s.

You might be wondering why. It also depends on the backline. When I'm subbing and the band isn't playing in-ear but with FRFR wedges or traditional cabinets, I use the ew100 because it has no latency.

With In Ear or in small clubs where the speaker is very close, I use the LIne 6 G70. This is the transmitter with the lowest latency of 1.5 ms (Line6 says so and I can confirm it with measurements)
But you can't use it in big shows with over 300 guests. As it transmits in the WLAN range, there are dropouts. In this case I use the Sennheiser ew-d. Very stable and only 1.8 ms latency. By the way. For In Ear I use the ew G4 from Sennheiser. IEM transmitter are not digital for a good reason ;-)

Any transmitter that adds more than 4 ms latency is nonsense in my opinion. So many Shure, Sennheiser transmitters that transmit in the WLAN range. Xvive is absolutely terrible in this respect.

The same applies to those that add several digital pedals to the FM9 loop. I have already seen pedalboards with 3 Eventide H9 pedals. And each one adds latency. In addition to the latency of the algorithms, you also have the latency of the DA-AD conversion.

The latency trap is lurking everywhere. Two years ago - I had just rebuilt my studio - I got a visit from my good buddy Thomas Blug from Bluguitar. I had just received the first FM9 and he had his AMP X prototype with him. So we connected both devices via my RME Fireface and Thomas started to play the FM9. He stopped immediately and said that the latency was too high and it didn't feel good. I then played and had to agree with him. Funnily enough, we also felt latency on his Amp X. Eventually we found out that my KS digital speakers (absolutely fantastic speakers by the way) had added 6 ms of latency. Then there was the latency of the conversion via the RME. So we stood 2 metres away from the speakers and in the end it added up to 10 ms.
The next day I sold my KS digital again and bought PSI Audio A21 speakers. Unbelievable linear speakers with an elaborate analogue crossover and phase correction. 0 ms latency.

And that's what I also like about the REDSOUND speakers. Great sound, very high-quality speakers and power amplifiers and an analogue crossover. No DSP. Okay, in my opinion the crossover could be designed even better, but that's a question of money. The design of an analogue crossover is so time-consuming that many users don't want to pay the extra price. Redosunds are the best choice for me at the moment. Apart from Meyer Sound. But unfortunately I'm not a millionaire.

So if this customer has realised that the latency is too high for them, this is important feedback. I'm sure that the non-resting developers at FAS are already working on an optimisation. And if they win one ms, that would be great. Then you can add something elsewhere ;-) And if not, that's good too. Because the sound is important. And it's fantastic. As Guthrie Govan can probably confirm ;-)
 
Last edited:
Hi, last week I received my Fm9 turbo and I noticed a bit of latency. It was odd because having owned the Axe FX standard many years ago and the Kemper more recently, I never cared about the latency those units had, so I assumed Fm9 would be on par if not better. I tested it and I measured a 4.6 latency on the Fm9. The amp block is adding almost 3ms of latency. The guys at G66 tested their units with betters tools and confirmed my findings, they also said that this must be a firmware thing and it will be fixed soon.

Thank you!
Until Fractal chimes in I won't believe anything is firmware related.
G66 is a reseller, not a co-developer.
 
Last edited:
Hi everyone! Thanks for taking the time of sharing your thoughts about my message. Let me try to explain the issue more accurately, why it bothers me and how I tested it. Please excuse my english cause I'm not native.

I been playing guitar for 20 years now. I work as a musician / producer / mixing engineer. I've used many gear over the years, tube and solid state amps, modelers (Pod, Axe FX Standard, Kemper), plugins... I understand that any digital gear has latency because the signal have to go to A/D, the actual processing of the signal, D/A, etc. I've just sold a kemper, which had served me well over the last 6 years. It's latency never bothered me. I just didn't thought of it.

When Fm9 Turbo showed up, It didn't feel right. NO, it was not unplayable and in a live situation, I probably didn't noticed as much. But in the studio it felt a little bit disconnected. Like playing thru a plugin. Not what I was expecting from probably the top floor modeller in the world.

So, trying to figure out if something was wrong with my unit, I tested it.

How i did it?

- First i measured the latency round-trip of my converters (Metric Halo ULN-8 mkIV) so I could took this off the equation.

- I measured too the latency added with the reamp box I was going to use to send the signal (which was negligible).

- For testing purposes I used a very short sinewave audio file, recorded the file going thru different units and comparing the results with the original wave position.

- I tested the kemper with a simple patch, I tested a real amp thru the Universal Audio OX box and I tested the FM9 (Just one amp with one cabinet, nothing more)

- Fastest was the Real amp with the OX at less than 3ms, then the Kemper just over 3ms and then the FM9 at almost 5ms.

- I tested the Fm9 with and empty grid and found that the amp block was adding almost 3ms of latency.

Now, I know these measures aren't gonna be super accurate because my daw is not the best tool to measure this with hight precision, but it gets you in the ballpark.

Now, I do not care about numbers, but I do care about how it feels, and right now my Fm9 do not feel great. (it sounds amazing dough)

I reported to G66, because Fractal says in their website that you should contact the seller if you didn't bought direct.

Their tech-support guy, answered to me that he will test it properly and so he did. He tested the Axe FX III, the FM3 and the FM9 with an oscilloscope and he said that to his surprise, the Fm9 had 4.6ms of latency. In comparison, the Fm3 had 2.9ms and the Axe FX III had 2.2ms in quality mode.

In my book it makes no sense that the FM9 is slower than the FM3, if anything shoud be faster because it has more processing power.

That's why the G66 guy thinks it is a firmware issue, and it will be fixed. I guess this little difference must be meaningful to me, otherwise I wouldn't have noticed.

Anyway, thats the whole story.

Thanks for listening!

Kind regards.
 
Hi everyone! Thanks for taking the time of sharing your thoughts about my message. Let me try to explain the issue more accurately, why it bothers me and how I tested it. Please excuse my english cause I'm not native.

I been playing guitar for 20 years now. I work as a musician / producer / mixing engineer. I've used many gear over the years, tube and solid state amps, modelers (Pod, Axe FX Standard, Kemper), plugins... I understand that any digital gear has latency because the signal have to go to A/D, the actual processing of the signal, D/A, etc. I've just sold a kemper, which had served me well over the last 6 years. It's latency never bothered me. I just didn't thought of it.

When Fm9 Turbo showed up, It didn't feel right. NO, it was not unplayable and in a live situation, I probably didn't noticed as much. But in the studio it felt a little bit disconnected. Like playing thru a plugin. Not what I was expecting from probably the top floor modeller in the world.

So, trying to figure out if something was wrong with my unit, I tested it.

How i did it?

- First i measured the latency round-trip of my converters (Metric Halo ULN-8 mkIV) so I could took this off the equation.

- I measured too the latency added with the reamp box I was going to use to send the signal (which was negligible).

- For testing purposes I used a very short sinewave audio file, recorded the file going thru different units and comparing the results with the original wave position.

- I tested the kemper with a simple patch, I tested a real amp thru the Universal Audio OX box and I tested the FM9 (Just one amp with one cabinet, nothing more)

- Fastest was the Real amp with the OX at less than 3ms, then the Kemper just over 3ms and then the FM9 at almost 5ms.

- I tested the Fm9 with and empty grid and found that the amp block was adding almost 3ms of latency.

Now, I know these measures aren't gonna be super accurate because my daw is not the best tool to measure this with hight precision, but it gets you in the ballpark.

Now, I do not care about numbers, but I do care about how it feels, and right now my Fm9 do not feel great. (it sounds amazing dough)

I reported to G66, because Fractal says in their website that you should contact the seller if you didn't bought direct.

Their tech-support guy, answered to me that he will test it properly and so he did. He tested the Axe FX III, the FM3 and the FM9 with an oscilloscope and he said that to his surprise, the Fm9 had 4.6ms of latency. In comparison, the Fm3 had 2.9ms and the Axe FX III had 2.2ms in quality mode.

In my book it makes no sense that the FM9 is slower than the FM3, if anything shoud be faster because it has more processing power.

That's why the G66 guy thinks it is a firmware issue, and it will be fixed. I guess this little difference must be meaningful to me, otherwise I wouldn't have noticed.

Anyway, thats the whole story.

Thanks for listening!

Kind regards.
I don't typically like to tag Cliff, but the difference between FM3 and FM9 does seem odd.

The processors in the FM3 and FM9 are the same - but the FM9 has 2 of them. The Amp modeling runs on one core, so I would expect that they would be the same or very close unless the FM9 has better modeling - which could be.

@FractalAudio - does this sound right?
 
I don't typically like to tag Cliff, but the difference between FM3 and FM9 does seem odd.

The processors in the FM3 and FM9 are the same - but the FM9 has 2 of them. The Amp modeling runs on one core, so I would expect that they would be the same or very close unless the FM9 has better modeling - which could be.

@FractalAudio - does this sound right?

I wonder if this has anything to do with the way processing may be split across processors rather than using only multi-processing. May not have used precisely the correct terminology there, but I have seen other modelers where the "handoff" from one set of cores to another incurs a latency penalty.

That, and maybe the potential for as much as twice the processing needing to complete, much of it sequential in nature - passed from block to block.
 
Hoping for that FW6 beta for the holiday weekend!
What makes you think they'd make such a big jump from 5.1 to 6.0 beta? I was thinking more like 5.2 to fix out any existing bugs. FW 6.0 would have to consist of some major changes to be a monumentus upgrade and that doesn't happen overnight.
 
Back
Top Bottom