Adventures in Accuracy

[aens]

The problem is that you have to use fender’s inbuilt IRs for a fair comparison, as the power amp modelling is inbuilt there. This is something that fender are going to change, according to the Facebook page.

To be honest, fender TMPs fender amp models generally sound a hair better than my fm3 turbo’s when you compare it that way (at least to my ears). The non-fender amps are not as good though. Also, fractal have more desirable fender amps available, currently.

Did you export the Fender IR’s to Fractal and compare them? If you didn’t, it’s not a fair comparison either.

I’m not trying to say TMP models can’t sound good and sure they can. The thing is TMP models are not yet accurate and have been ”sweetened” to sound good with other techniques (like the ones Cliff mentioned). Comparing that to a product that is replicating real life counterparts in the finest detail no matter if they sound good or bad (Fractal) is kind of comparing apples to oranges. Of course one can say their tweaked to death Line 6 Pod Fender preset sounds better than a Fractal preset or even the real amp, but it has nothing to do with authenticity.

 

[Daveporter33765]

The fairest comparison between one modeler and another is to use just the amp models, cab IRs and effects provided as stock with each modeler and develop presets that model a physical brand and model of amp that the majority of guitarists have used and have experience with…like a Fender Deluxe black face. Then the results will speak for themselves. I have only had one experience with the Fender TMP, and it was limited, however, I, and the units owner, found that we were not able to easily dial-in an acceptable Fender Deluxe Reverb preset. If any modeler should be capable of producing quality, accurate Deluxe Reverb tones, it should be a Fender modeler. As I said in my earlier post, neither the owner nor I had any experience with this modeler and we had not read the manual at that point. But, in the past, I have always been able to dial in an acceptable preset for standard amp models with limited experience the first time I encountered other modelers. So, for me, and this is only my opinion, the Fender TMP is not ready for prime time.

UPDATE. I just spoke to my friend to see if he had made any progress in developing presets on his TMP since his last band practice. He told me he gave up and returned the unit. He currently owns a Headrush Pedalboard and is going to purchase a FM9 to replace it now that he has tried the TMP and did not care for it.

 

[Frodebro]

The fairest comparison between one modeler and another is to use just the amp models, cab IRs and effects provided as stock with each modeler and develop presets that model a physical brand and model of amp that the majority of guitarists have used and have experience with…like a Fender Deluxe black face. Then the results will speak for themselves. I have only had one experience with the Fender TMP, and it was limited, however, I, and the units owner, found that we were not able to easily dial-in an acceptable Fender Deluxe Reverb preset. If any modeler should be capable of producing quality, accurate Deluxe Reverb tones, it should be a Fender modeler. As I said in my earlier post, neither the owner nor I had any experience with this modeler and we had not read the manual at that point. But, in the past, I have always been able to dial in an acceptable preset for standard amp models with limited experience the first time I encountered other modelers. So, for me, and this is only my opinion, the Fender TMP is not ready for prime time.

UPDATE. I just spoke to my friend to see if he had made any progress in developing presets on his TMP since his last band practice. He told me he gave up and returned the unit. He currently owns a Headrush Pedalboard and is going to purchase a FM9 to replace it now that he has tried the TMP and did not care for it.

I think Fender put themselves in a really tough spot bringing the TMP to market at that price point. If things were more fully developed they may have fared better, but to charge the same amount as an FM9 sets them up to be judged on that level, which I don’t think they were expecting (and definitely weren’t prepared for).

I think their aim was to market a modeler for people who otherwise avoided them (which they had great success with in their Tonemaster combos), but they drew the attention of the hardcore, experienced users who are used to a certain level of modeling and features.

Don‘t get me wrong, I do want them to be successful because they have the brand recognition to bring a ton of new people into the digital world (which means more sales down the road for Fractal, Line6, etc).

 

[Daveporter33765]

I think Fender put themselves in a really tough spot bringing the TMP to market at that price point. If things were more fully developed they may have fared better, but to charge the same amount as an FM9 sets them up to be judged on that level, which I don’t think they were expecting (and definitely weren’t prepared for).

I think their aim was to market a modeler for people who otherwise avoided them (which they had great success with in their Tonemaster combos), but they drew the attention of the hardcore, experienced users who are used to a certain level of modeling and features.

Don‘t get me wrong, I do want them to be successful because they have the brand recognition to bring a ton of new people into the digital world (which means more sales down the road for Fractal, Line6, etc).

Fender, who is owned and controlled by corporate interests with bean-counters who set the goals and objectives for the company, is at a serious disadvantage because of that relationship. Fractal is owned and controlled by engineers who develop products based on their love of the technology and the impact their products have on their users who love music making as much as they do. Bean-counters are not devoted to any product beyond its ability to generate sales and profits. Therefore, low production cost, “good enough“, high margin products are their goal…not the production of high-quality, great value, long-lived, user loved products that are reasonably priced (possibly generating lower margins). To bean counters, high profits and increasing stock values are the goal. So, it is going to be interesting to see how far Fender’s bean counters will allow the engineers to go in trying to improve the TMP and compete in the market…they have already forced the cost of the units to be way too high as you so correctly stated above.

Fender is far from the only modeler company that is experiencing these problems, most companies controlled by large corporations are in the same boat. Fractal Audio is an anomaly in the industry; we the Fractal customers, are surely fortunate indeed!

 

[Frodebro]

Fender, who is owned and controlled by corporate interests with bean-counters who set the goals and objectives for the company, is at a serious disadvantage because of that relationship. Fractal is owned and controlled by engineers who develop products based on their love of the technology and the impact their products have on their users who love music making as much as they do. Bean-counters are not devoted to any product beyond its ability to generate sales and profits. Therefore, low production cost, “good enough“, high margin products are their goal…not the production of high-quality, great value, long-lived, user loved products that are reasonably priced (possibly generating lower margins). To bean counters, high profits and increasing stock values are the goal. So, it is going to be interesting to see how far Fender’s bean counters will allow the engineers to go in trying to improve the TMP and compete in the market…they have already forced the cost of the units to be way too high as you so correctly stated above.

Fender is far from the only modeler company that is experiencing these problems, most companies controlled by large corporations are in the same boat. Fractal Audio is an anomaly in the industry; we the Fractal customers, are surely fortunate indeed!

The corporate bean counter mentality is going to hurt Fender in the long run as far as their modeling products are concerned. Line 6, although owned by YGG, is (so far) treated with a hands-off approach and their team seems to have quite a bit of freedom to do their own thing. With Fractal, obviously Cliff is the owner and primary mover & shaker, and is very accessible to answer user questions and quick to jump on bugs as well as improvements.

At this level in the modeling market, users are simply used to the depth of communication and activity that we receive from Fractal and Line 6. The Fender reps are doing their best, but they just aren’t in a position to get things done with the speed at which their competitors are running at.

 

[aens]

The fairest comparison between one modeler and another is to use just the amp models, cab IRs and effects provided as stock with each modeler and develop presets that model a physical brand and model of amp that the majority of guitarists have used and have experience with…like a Fender Deluxe black face. Then the results will speak for themselves.

I disagree with you, especially when some manufacturers use EQ’d or smoothened IRs. If you throw a different IR to a modeler that doesn’t have authentic amp modeling it should not sound that good. By comparing two products with different IR’s you are just comparing two sounds together, not two modelers. That’s why the IR should be the same.

 

[Daveporter33765]

I disagree with you, especially when some manufacturers use EQ’d or smoothened IRs. If you throw a different IR to a modeler that doesn’t have authentic amp modeling it should not sound that good. By comparing two products with different IR’s you are just comparing two sounds together, not two modelers. That’s why the IR should be the same.

If one wishes to compare modelers, then one must utilize the capabilities of each modeler as it comes from the developer. Many users will not mix and match IRs from different producers, but use the IRs provided with the modeler itself.

if the TMP, or any modeler, can not produce quality, competitive tones as other modelers as it comes when purchased, then it is deficient and not up to the task.

What you are proposing would be equivalent to purchasing two autos and then claiming that they could not be compared one against the other unless the brand new tires provided on each car were first removed and new high performance tires were placed on both cars before the comparison. In reality, most new car buyers would drive The new car off the lot with the tires it came from the factory with and compare it to his/her neighbor’s new car with the tires it came with.

What makes Fractal units so exceptional is that they are great as they come from the factory and they improve as new free FW updates are released frequently throughout the unit’s long useful life!

 

[aens]

If one wishes to compare modelers, then one must utilize the capabilities of eachmodeler as it comes from the developer. Ma y users will not mix and match IRs from different producers, but use the IRs provided with the modeler itself.

if the TMP, or any modeler, can not produce quality, competitive tones as other modelers as it comes when purchased, then it is deficient and not up to the task.

Just a friendly reminder that this thread is about authenticity and accuracy

If you are comparing two amp modelers with two different IR’s, you are mostly comparing two miced up sounds and trying to decide which miced up sound you like better.

 

[Daveporter33765]

Just a friendly reminder that this thread is about authenticity and accuracy

If you are comparing two sounds with two different IR’s, you are mostly comparing two miced up sounds and trying to decide which miced up sound you like better.

That was exactly my point. It is the accuracy of the modeler as a whole this thread is about. If you mix and match IRs not present initially and natively provided with each modeler, then you are not comparing the accuracy of each modeler. I believe you proved your own argument to be invalid.

We obviously do not agree on how the difference in accuracy between two modelers should be measured. We appear to have two different views; Those reading our posts are free to consider each view. I wish you the best. Have a great day.

 

[aens]

That was exactly my point. It is the accuracy of the modeler as a whole this thread is about. If you mix and match IRs not present initially and natively provided with each modeler, then you are not comparing the accuracy of each modeler. I believe you proved your own argument to be invalid.

We obviously do not agree on how the difference in accuracy between two modelers should be measured. We appear to have to different views; those reading our posts are free to consider each view. I wish you the best. Have a great day.

No offence taken at all. If you use the same IR between different units, then the comparison is about the amp modeling being accurate or not. This includes the interaction between the amp model and the speaker.

If you use two different products, same amp model but use a different IR, you are mostly comparing two sounds together and it doesn’t have much to do with accuracy of the modeling. More like matter of taste trying to decide which sound do you like more and figure out which one is more accurate with many different variables in play. Measuring the accuracy is not possible, because the impulse response is different. Therefore it becomes a matter of taste. Let alone trying to compare a real amp with a miced up speaker to two modelers using different IR’s. They can all sound widely different.

Different impulse responses have many variables: speakers sound different, mic placement sounds different, mics sound different, power amps sound different, preamps sound different, rooms sound different, different capturing methods sound different, heck even the cables can sound different.

 

[Frodebro]

If one wishes to compare modelers, then one must utilize the capabilities of each modeler as it comes from the developer. Ma y users will not mix and match IRs from different producers, but use the IRs provided with the modeler itself.

Not worth a poll, but I would bet that you are likely to see a notable percentage of Fractal users running third party IRs. If that is a normal trend for this particular level of digital user (which I bet it is), then being almost forced to stay in the box is something that many people would have a problem with. Note that one of the selling points on the Axe 3 is the huge amount of space available for user IRs.

 

[Daveporter33765]

Not worth a poll, but I would bet that you are likely to see a notable percentage of Fractal users running third party IRs. If that is a normal trend for this particular level of digital user (which I bet it is), then being almost forced to stay in the box is something that many people would have a problem with. Note that one of the selling points on the Axe 3 is the huge amount of space available for user IRs.

I am sure you are correct. Many advanced users add IRs to their units. My point was not that others do not use additional IRS, but that if one wants to compare the accuracy and quality of one modeler against another modeler using only the IRS supplied with each modeler as stock along wi/th only the amp models and effects supplied as stock in each unit would allow for a comparison of each unit's capability to accurately model and produce tones of specific well known physical amp models. If third party IRs are used each modeler is no longer being compared against thei other because the third party IRs would not be present on stock units.

 

[Frodebro]

I am sure you are correct. Many advanced users add IRs to their units. My point was not that others do not use additional IRS, but that if one wants to compare the accuracy and quality of one modeler against another modeler using only the IRS supplied with each modeler as stock along wi/th only the amp models and effects supplied as stock in each unit would allow for a comparison of each unit's capability to accurately model and produce tones of specific well known physical amp models. If third party IRs are used each modeler is no longer being compared against thei other because the third party IRs would not be present on stock units.

Thing is, using the same IR when comparing models eliminates any variables, so what you are comparing is specifically amp model “A” to amp model “B.”

 

[Daveporter33765]

Thing is, using the same IR when comparing models eliminates any variables, so what you are comparing is specifically amp model “A” to amp model “B.”

I was comparing Modeler A against Modeler B…not just two particular amp models on their own against each other. This is because there is much more that goes into determining the accuracy of a a model provided with any particular modeler…it’s not just the amp algorithm in isolation on its own. So, if one want to compare a physical Fender Deluxe Reverb combo amplifier “system” against Modeler A and Modeler B one would have to use each modeler’s components including the on-board IRs, which have been designed to work together as a “system”, just like the physical amp.

 

[Frodebro]

I was comparing Modeler A against Modeler B…not just two particular amp models on their own against each other. This is because there is much more that goes into determining the accuracy of a a model provided with any particular modeler…it’s not just the amp algorithm in isolation on its own. So, if one want to compare a physical Fender Deluxe Reverb combo amplifier “system” against Modeler A and Modeler B one would have to use each modeler’s components including the on-board IRs, which have been designed to work together as a “system”, just like the physical amp.

I get the reasoning behind it, but that doesn’t change the fact that power users just love to go down the IR rabbit hole, many roll their own, and telling them that their carefully curated library of 14,000 IRs will not work as well as the 20 factory IRs that came in the unit, well, we all know how well that is being received.

 

[Daveporter33765]

I get the reasoning behind it, but that doesn’t change the fact that power users just love to go down the IR rabbit hole, many roll their own, and telling them that their carefully curated library of 14,000 IRs will not work as well as the 20 factory IRs that came in the unit, well, we all know how well that is being received.

i am not being clear. I am not stating that using third-party IRs is not desirable or common, in fact I totally agree. I am talking about comparing two modelers against each other to determine how well each unit is able to produce accurate renditions Of various models. During that comparison, when one modeler is being compared to the other, to use critical components that are not provided with each modeler could provide results that do not accurately compare the two modelers. Using third-party IRs when comparing the two modelers, could easily result with inaccurate results of the comparison.

We have beat this issue to death. The bottom line is that, in my limited experience with a Fender TMP modeler, the tones that can be obtained with a stock unit do not compare to those provided by my Axe Fx modeler…not even close. Perhaps Fender will provide upgrades that will improve its performance and the quality.

Have a great day and happy music making.

 

[MicroMort]

I think Fender put themselves in a really tough spot bringing the TMP to market at that price point.

This. the TMP is a 400-600 dollar modeler at this time. Maybe with updates they can bring it up to be worthy of the price they are asking for. By the same line of thinking, the FM9 is probably the best bang for your buck in the fractal lineup.

 

[Daveporter33765]

This. the TMP is a 400-600 dollar modeler at this time. Maybe with updates they can bring it up to be worthy of the price they are asking for. By the same line of thinking, the FM9 is probably the best bang for your buck in the fractal lineup.

I agree with you. The Fender Tone Master Pro’s current price in the US is $1,699. At that price, a Fractal FM9 can be purchased. For $100 less. In my opinion, the choice should be obvious.

Have a great week!

 

[citadel]

Did you export the Fender IR’s to Fractal and compare them? If you didn’t, it’s not a fair comparison either.

I’m not trying to say TMP models can’t sound good and sure they can. The thing is TMP models are not yet accurate and have been ”sweetened” to sound good with other techniques (like the ones Cliff mentioned). Comparing that to a product that is replicating real life counterparts in the finest detail no matter if they sound good or bad (Fractal) is kind of comparing apples to oranges. Of course one can say their tweaked to death Line 6 Pod Fender preset sounds better than a Fractal preset or even the real amp, but it has nothing to do with authenticity.

Absolutely but you can’t do that because you would have to turn off specific portions of the power amp modelling for the fractal, which I don’t think could be done (not easily, at least). It’s not the entire power amp, from my understanding, just the interactions with speakers. I don’t think you can export the inbuilt IRs from the TMP also.

The point is also that you don’t have to tweak the TMP to “death”, ala your line 6 example. It’s there from the go. So I am talking about quick “out of the box” sounds.

The modelling approach is very much like a black box these days. No one is reverse engineering Deep learning networks…which most people are using, even for component-wise modelling.

My point is that the apples to apples test has not been done and that the only fair test you can probably do now is the “out of the box” quick tests, until fender fix their IR architecture.

 

[citadel]

I agree with you. The Fender Tone Master Pro’s current price in the US is $1,699. At that price, a Fractal FM9 can be purchased. For $100 less. In my opinion, the choice should be obvious.

Have a great week!

The TMP is $2299 AUD in Australia versus $3299 AUD for the FM9. Fractal is too expensive here, unfortunately.