A bit about flattening the frequency characteristics and easier work with the sound (long)

[Lecchu]

Many of us know these curves:

These are equal loudness curves for different sound levels. The so-called Fletcher-Munson curves (perception of human hearing). The effect associated with them causes, among other things, that the preset prepared in Fractal in studio conditions at a volume of e.g. 50-60dB will sound different under the conditions of concert sound system 100dB (for now I ignore the influence of various speakers in the satudio and on the concert - about this below) - in conditions high sound level we will usually "miss" the middle tones. This is of course due to the fact that for 50dB in the preset studio we will set the midrange will be about -15 to even -20dB "weaker" than the low and high tones. Whereas for 100dB this difference should sleep up to 10dB for high tones (and 0dB for low tones) - see the diagram above. So the sound set in the studio (preset) for 50dB will sound dramatically different at high volume.
Of course, we can improve the preset so that it sounds good in high volume conditions (boost the middle tones e.g. with GlobalEQ in the Fractal settings). But it is often difficult. There is often no time before the concert to fine tune the preset (especially if the concert is festival, there are many bands and you are not its star; D). Anyway, even having time to tune the preset on high volume is not easy and comfortable, because after a while high sound level causes "headache" and loss of perception of sound details.

Therefore, wouldn't FAS consider adding functionality to its products (rather than a block, only a position in global options, something similar to GlobalEQ) to reduce the impact of the Fletcher-Munson effect. I imagine that in connection with a particular output (Outpu1, Output 2, etc) we have a value associated with the target sound level. Then in the studio (at home) we define the starting sound level ("start level" - the sound level at which we create and tune the preset), while elsewhere (in the rehearsal room, during the concert = "target level") we set at one point only as far as the level of sound increases / decreases in a given location (+ 10dB, + 20dB, + 40dB, -10dB for the bedroom, ect). And then the Fractal software would "apply" to our preset so many changes in individual sound bands to try to minimize the Fletcher-Munson effect (i.e. convolution of the preset signal with something like an impulse response that is different between the FM curve for the "target level" "a" start level "). This functionality will probably not solve the whole problem related to changing the volume, but it will allow you to tune the sound much faster, e.g. before the concert in a new place and at a new volume.

Of course, a certain problem will be determining the "start level" value (at home), it would be ideal to have a level meter. But even so, setting it to a Sound Meter app on your smartphone will give you a pretty good initial point.

 

[Lecchu]

The second topic that comes to my mind is whether FAS products could "improve" the frequency response of the target speakers (physical sound system).
Let's look at this chart:

This is a graph showing the frequency response of an example Celestion F12-X200 speaker (here: link). The red line was formed after examining by the manufacturer how the speaker reacts to individual frequencies, the green horizontal straight I drew it.
My question is whether FAS would add another functionality to its products, which would prepare the output signal in such a way as to reduce the impact of the physical speaker (which will not always be FRFR). Mathematically, this would probably be the convolution of the signal from the preset and "IR" being the difference between the green and red lines (digital filter direct form I?). For the above loudspeaker, before leaving Fractal, we would weaken 2kHz by about 7dB, 3kHz weaken by about 4dB, for 6kHz we would add about 4dB etc. And playing a physical speaker alone will add 7dB for 2kHz, 4dB for 3kHz - and ultimately we would have a flatter characteristic. The only thing you would need is the above red curve in numerical form (not only as a graphic) and transform it into a form understandable by Fractal. I think that with the extension of such functionality, the loudspeaker / cabs manufacturers would provide this type of data in a form that can be downloaded to Fractal (or users have entered such data themselves).

What's the point? Not everyone has FRFR speakers (e.g. in Europe it is not possible to buy CLR NEO from Atomica at present). And adding such an "IR flatten for a loudspeaker" could allow you to enjoy better sound for a wider range of" worse "loudspeakers/cabs.
This functionality should probably be connected to individual and outputs (Outpu1, Output2, etc) - something like GlobalEQ. Rather, this should not be a preset feature. After insert several such curves to the Fractal (one for the home speakers, another for the studio speakers, rehearsal room, concert loudspeakers, etc), it would only be possible to change these curves in the Fractal's global settings to eliminate the influence of different physical loudspeakers on the preset's sound.

The default / start state of such a module would be of course the 0dB level (or "off"). Only when you have / acquired the appropriate "flattening" IR could you upload and use it.

The combination of both functions (the nulling of the Fletcher-Munson effect and the non-influence of the target columns) may reduce the need for continuous tuning of the prestige sound in each place and at different loudness.

 

[lqdsnddist]

You’ve got it a little reversed, the human auditory system has the best perception for the mid frequencies. The curves represent minimum threshold levels; ie, we can perceive a mid freq tone at a lower intensity level than we can extreme highs and lows

 

[Lecchu]

"human auditory system has the best perception for the mid frequencies"
Yes, on typical sound level 50-60dB mids are 20db bellow lows and highs to hear "equal volume" (look on first graph). It just means that we are ready to hear mids - just means that 20dB will be quieter than low and high tones, and we will already perceive them at "50dB" (in this case lows need 70dB to be perception as "50dB")

Therefore, we will prepare the preset at the volume level so that all ranges hear similarly, so mids will be about 20db quieter than low and high (we have 70dB lows, 50dB mids, 65dB highs).

And now you set output level (from phisicall cabs - that means you turn up power on phisicall amps) to 100dB. The graph shows that to hear everything similarly, the mids should be on the same level as the low and only 10dB quieter than the high.
But you prepared the preset, that mids are -20dB (in signal in audio path) in comparison to lows (and highs). So after turning up to 100dB, they will be missing for this reason.

 

[Pwrmac7600]

OR, you could dial in your presets at rehearsal where you are playing at gig levels, (like you would with a real amp and cab) then save that as a live version of your preset, so you could have two version of your presets, live and studio.

 

[2112]

This has been thrashed to death over the years, but there's no substitute for dialling in your sounds for the context you intend to use them in. Using features like global blocks (a global PEQ for live presets for example) and the perform screens can make tweaking things live easier.

I'd hazard a guess that this would be faster and more intuitive than setting up an external mic and shooting a reference IR of every playback system you ever use.

 

[Lecchu]

OR, you could dial in your presets at rehearsal where you are playing at gig levels, (like you would with a real amp and cab) then save that as a live version of your preset, so you could have two version of your presets, live and studio.

Yes, of course you can do that - but you don't always have time before the concert to play settings in AXE-Edit, you don't always stand next to the mixer to listen to what's going on, and in the club at high volume you have little time before high sound will not cause that the ears will fade.
The use of FM curves would simply simplify this process (maybe even only to change one parameter - the target sound level, that will be ideal ;D)

 

[Lecchu]

This has been thrashed to death over the years, but there's no substitute for dialling in your sounds for the context you intend to use them in. Using features like global blocks (a global PEQ for live presets for example) and the perform screens can make tweaking things live easier.

PEQ (or GlobalEQ) propably will be still needed. But use "FM convolution curves function" imo can do all sound tweking for different level mutch easier.

I'd hazard a guess that this would be faster and more intuitive than setting up an external mic and shooting a reference IR of every playback system you ever use.

No external mic, no shooting reference IR on every systems - IR for speaker as file only from a loudspeaker manufacturer. Loudspeaker manufacturer do graphes with freqence respone of speakers for specifications/marketing folders, therefore he has this data in numeric form.
If for a given system we do not have the right IR from the manufacturer - well, we will be tuning the sound "the old way" as now

 

[biskitboy]

"human auditory system has the best perception for the mid frequencies"
Yes, on typical sound level 50-60dB mids are 20db bellow lows and highs to hear "equal volume" (look on first graph). It just means that we are ready to hear mids - just means that 20dB will be quieter than low and high tones, and we will already perceive them at "50dB" (in this case lows need 70dB to be perception as "50dB")

Just want you to know that you are trying to explain this to someone who has a doctorate in audiology @lqdsnddist.

 

[Lecchu]

Just want you to know that you are trying to explain this to someone who has a doctorate in audiology @lqdsnddist.

I'm very glad @lqdsnddist has a PhD in audiology, good for him.
It is possible that unintelligible creeping in somewhere (English is not my native language).

However, I will defend the thesis that the preset prepared "for the ears" for 50dB, boosted without any changes, on an audio system identically carrying the sound also at 100dB will sound like we think that it is too high and low (which in other words means that has too few mids) - this is what appears in FM graphs.

 

[biskitboy]

Thanks @Lecchu. There probably is a little bit of a language barrier here, but that is totally ok. This has been rehashed on this forum many times over the years, and there are just a lot of different issues with this. That doesn't mean there should or should not be some type of compensation. It just won't be perfect by any means. If it existed, then people will want that tweaked and we are back to having global eq/peq blocks. We already have that feature.

 

[Pwrmac7600]

Yes, of course you can do that - but you don't always have time before the concert to play settings in AXE-Edit, you don't always stand next to the mixer to listen to what's going on, and in the club at high volume you have little time before high sound will not cause that the ears will fade.
The use of FM curves would simply simplify this process (maybe even only to change one parameter - the target sound level, that will be ideal ;D)

Yes the point I was trying to make is that when i gigged with a tube amp and cab, I never had time to do that either. I would dial in my rig at rehearsal at stage volume, and then use those same settings for the gig, maybe slightly adjusting treble/mid/bass to suite the room/stage.

Any pre made settings that could be implemented would not carry from room to room anyway as each room is extremely different. So although you may "in theory" compensate for FM, the implementation would only work "in reality" if you played in the exact same room over and over again. different rooms may have inherent issues such as being a boomy room that exaggerates low end, or a dull room, or an overly bright room. which will then cause your pre programmed FM curves to be useless. So unless you can analyze every room you play in ahead of time it would be a mute point.

 

[Lecchu]

I am not saying that the use of "differential FM curves" and / or null influences of speakers through the use of "negative" IRs will solve the problem. It is obviously not, because the room, cubature, open air, etc. are still important. However, I think the setting would be much easier:

• you come somewhere and you have to play 40dB louder than in "home". You enter GlobalFMCurvesFunction in Fractal, switch to + 40dB, the firmware already captivates some of the problems resulting from the FM effect. For all presets
• you come to this place with your sound system. At home, you set the preset to HS8 - at that time you had the "negative IR8 for HS8" fastened to zero the effect of this speaker when setting the preset. Now you are in a small club room, you have two columns with Celestiron FX200 with you (they are yours, but you do not play at home because you have HS8 at home). You enter the global settings of the Fractal, you switch the negative IR speaker from HS8 to CelestionFX200. The effect of Celestion speaker colorization is largely offset (as the HS8 colorization effect was turned off at home when creating a preset). For all presets
• you come to the rehearsal rooms; you play there at this Celestion, but more quietly, about 20dB louder than at home (and -20dB from what you did during the concert). You connect the negative IR for Celestion, the FM differential curve at + 20dB compared to what is at home.

These are pretty quick changes, right? Two shifts and we already have a fast and good point to further refine - with the help of globalEQ (or PEQ as a global block). Or maybe nothing will have to be done anymore?
And it will work correctly (the same way) if you change something at home in the preset - during a concert or rehearsal of changes they will "update" themselves; D
You can of course have separate presets for the home, and separate for the rehearsal room (and more for the concert). But if you change something in some "home" preset - now you have to make these changes in the presets from the rehearsal or concert (but you will not make these changes at home, because these are not the loudspeakers and the loudness).

 

[Reaper]

There is a wish for some sort of a snapshot function for the global EQ pages of output 1 and 2.
So you can easily - if I understand it correctly - swap pre-defined GEQ settings. i.e. for headphone, bedroom, a little loud small gig and a full blown loud open air stage.
Wouldn't this be a solution to Lecchu's problem?
At least a partial solution, 'cause you will have to pre-programm the GEQ snap shots. But like @2112 / Leon said: no situation is the same so automated F-M effect cancellation is an utopia.

(hope this all is linguistically correct because like Lecchu I'm not a native English speaker; me from Yek country )

 

[Lecchu]

Wouldn't this be a solution to Lecchu's problem?

Is it really me only who has this problem?
I thought everyone had to deal with this and I am looking for a path that will help everyone

Snapshots in GlobalEQ would also be pretty good, although GlobalEQ doesn't have the same frequency resolution as IR (if they had, we wouldn't need IR for CAB blocks, just a "terce" EQ )

 

[DLC86]

I think an option like this wouldn't be worth the effort cuz there are too much variables.

• Resonances on a room/venue will have a much bigger effect on the frequency response than any variance there might be between different brands of FRFR speakers.
• FM curve is different for each one of us.
• FM curve depends on the SPL and that means it also varies depending on the distance of the listener from the speaker.
• If you still want to make some sort of room/speaker correction I think it's better to do it for the PA speakers/mixer output

PS: Regarding the last point, a lot of sound guys do this simply by feeding some pink noise to the PA and then use a measurement and an RTA to find out problematic frequencies and correct them with a GEQ at the mixer out. That won't certainly take into account all the variables, but it's a quick way to correct the biggest issues and get you 90% there.

 

[Pwrmac7600]

Is it really me only who has this problem?

No don't think it's only you who has this problem, but I do think that treating the Axe FX like you would any actual amp in a live environment is still the easiest, and fastest solution.
Dial in at gig volume, tweak Treb/Mid/Bass quickly at gig to fine tune.

Never once when gigging with a actual amp did i say I wish i had a way of flipping a switch for studio and live usage. in the studio i would dial my amp in to suit the recording, and I always had my live settings marked in tape on my amp head, each gig that was where i set my dials, and when walking on stage I would bang a few chords and tweak the eq on the amp quickly before jumping into the first song. quick and easy.

I am not going to knock you for what you are asking for, believe me I get it. Just think sometimes as guitarists it is in our nature to overthink things.

 

[painkiller]

I usually dial in my presets at home with headphones playing along a playback. Later during band rehearsal at high volumes I reduce only bass and treble or presence (I usually don't touch the mids) until I like it. Nothing else. This works for me. If I find it necessary I use a bit of global EQ depending on the location.
These presets will sound really bad back on bedroom volumes but they are intended for high volumes!

 

[mr_fender]

The Fletcher-Munson curves are from 1933 and have since been updated. The current standard for equal loudness contours is ISO 226: 2003 revision

 

[App12]

Is it really me only who has this problem?
I thought everyone had to deal with this and I am looking for a path that will help everyone

Snapshots in GlobalEQ would also be pretty good, although GlobalEQ doesn't have the same frequency resolution as IR (if they had, we wouldn't need IR for CAB blocks, just a "terce" EQ )

I think this is a great idea!