Melda on the bench
Posted: Sun Feb 25, 2024 2:02 am
This is the successor to an earlier, quite enthusiastic review about Melda’s products. Let´s see how good they hold up to those advance laurels.
Several months have elapsed, since I bought MTC (MTurboComp), looking for the perfect *versatile* compressor. In a way I found it, but not quite due to some practical shortcomings, but this would bust the envelope here.
Melda’s most outstanding feature is their multi-parameter (MP) language. This is a method for constructing your own devices with the core modules your unit provides, which would be envelope processors, filters, modulators, harmonics generators and similar in the case of MTC. Other units come with a lesser or bigger package of those.
This feature is what distinguishes Melda’s units by far from their competitors. Unfortunately it appears incorporated very sloppy and hastingly, and some of their MPs’ visualizations don’t work as expected, have unexpected side-effects, or plain simply don’t work at all. This makes it impossible to successfully create a device.
I cross-checked with an evaluation version of Reaper, which is officially supported by them, to avoid all discussions going into that directions. It is easier to get quickly rid of a customer if the DAW is not officially supported.
I sent them a big document, in-depth structured, that would help to pin-point the problem, bundled with a reference device. If somebody had genuinely looked into that, there would be no excuse to not see the problems.
After days, sometimes weeks, a very friendly customer support comes back and promises that they will look into it, but even after a major revision release (16.11) those things have not changed. In fact I receive the impression, they have not even be tackled.
I believe that not very many people use the MP feature, so fixing it is on low priority in the lights of continuously dishing out new units, which makes me believe that nothing will improve. It also looks as if they had limited personnel and that those few are busy inventing new stuff.
They have a web presence on KVRaudio, but I have not noticed any staff member chiming in on severe complaints like that. There you find a few lonely users that share your problem for consolation, but no one knowledgeable. So this forum cannot be taken for serious.
Another problem that persists is their version of emulations of iconic compressors. Interestingly, nobody seems to complain about the quality of those, or how good they are in emulating, but many complain in unison about missing some very basic practical functions.
Like there is no official information about which those icons are. They take the standpoint that they indeed used icons as a starting point, but later strayed from those by making them better (whatever this “better” may mean), while on the same time making some unmistakable references to those icons by means of naming conventions.
You therefore end up with a handful of potentially powerful specialized (character) compressors that are neither fish nor meat, with no information about them – similar to the icons but deemed better. There is no information about those units, and the manuals from hardware icons won’t work either.
Again, there seems no desire to improve this situation except the mantra “use your ears”, which starts to create nausea in me, being the practical kill-all excuse for further confrontation.
Their claimed feature of being able to cycle through the models for comparison won’t work by a mile, because there are some huge level inconsistencies besides natural incompatibilities between, say a FET timing stack and an opto timing stack. I must say that their EZ screen front plate idea could have been a huge step towards meeting this goal, but it is ridden with problems and inconsistencies that have not been thought to an end. In fact this “feature” can easily turn into a bug because they have made some over-simplifications that are not documented.
Akin to writing a manual “tank-driving for noobs”.
Scrutinizing those models in pursuit of a deeper understanding, alongside with manuals for hardware icons from the internet, I found a lot of unexplainable behavior, which I believe will not be part of the original design. Yet they claim so.
A question that remains yet to be answered is how they managed to squeeze a feed-back design (which most of the units they have there are…) into a feed-forward architecture, because on such a unit you cannot reach huge attenuation ratios despite what the knob setting suggests by design.
The suitable mantra is “machine learned…”
You can theoretically knit your own design or amend existing designs and then upload them to the community, but, as above, none of that is of any use without meaningful documentation. Indeed, you are well advised not to use any undocumented public preset without extensive testing purely from that standpoint.
You cannot enter help files, the uploading mechanism’s help is of little use and no assistance was received from the forum or even a direct request to customer support.
Program help is nicely done, but rigid and very often ambiguous. Some terms have been chosen unwisely and obviously by somebody, whose native language is not English. Again, no additional information is available and suggestions were ignored. The videos help only to an extent. I have offered assistance but with no response.
This all could be a shining gem if it were not treated so step-motherly. I get the impression that their customer support does not know what to do with me. None of the subjects I brought up are unearthly. The few informations that can be found involve extensive detective work.
Make your own decision.
Several months have elapsed, since I bought MTC (MTurboComp), looking for the perfect *versatile* compressor. In a way I found it, but not quite due to some practical shortcomings, but this would bust the envelope here.
Melda’s most outstanding feature is their multi-parameter (MP) language. This is a method for constructing your own devices with the core modules your unit provides, which would be envelope processors, filters, modulators, harmonics generators and similar in the case of MTC. Other units come with a lesser or bigger package of those.
This feature is what distinguishes Melda’s units by far from their competitors. Unfortunately it appears incorporated very sloppy and hastingly, and some of their MPs’ visualizations don’t work as expected, have unexpected side-effects, or plain simply don’t work at all. This makes it impossible to successfully create a device.
I cross-checked with an evaluation version of Reaper, which is officially supported by them, to avoid all discussions going into that directions. It is easier to get quickly rid of a customer if the DAW is not officially supported.
I sent them a big document, in-depth structured, that would help to pin-point the problem, bundled with a reference device. If somebody had genuinely looked into that, there would be no excuse to not see the problems.
After days, sometimes weeks, a very friendly customer support comes back and promises that they will look into it, but even after a major revision release (16.11) those things have not changed. In fact I receive the impression, they have not even be tackled.
I believe that not very many people use the MP feature, so fixing it is on low priority in the lights of continuously dishing out new units, which makes me believe that nothing will improve. It also looks as if they had limited personnel and that those few are busy inventing new stuff.
They have a web presence on KVRaudio, but I have not noticed any staff member chiming in on severe complaints like that. There you find a few lonely users that share your problem for consolation, but no one knowledgeable. So this forum cannot be taken for serious.
Another problem that persists is their version of emulations of iconic compressors. Interestingly, nobody seems to complain about the quality of those, or how good they are in emulating, but many complain in unison about missing some very basic practical functions.
Like there is no official information about which those icons are. They take the standpoint that they indeed used icons as a starting point, but later strayed from those by making them better (whatever this “better” may mean), while on the same time making some unmistakable references to those icons by means of naming conventions.
You therefore end up with a handful of potentially powerful specialized (character) compressors that are neither fish nor meat, with no information about them – similar to the icons but deemed better. There is no information about those units, and the manuals from hardware icons won’t work either.
Again, there seems no desire to improve this situation except the mantra “use your ears”, which starts to create nausea in me, being the practical kill-all excuse for further confrontation.
Their claimed feature of being able to cycle through the models for comparison won’t work by a mile, because there are some huge level inconsistencies besides natural incompatibilities between, say a FET timing stack and an opto timing stack. I must say that their EZ screen front plate idea could have been a huge step towards meeting this goal, but it is ridden with problems and inconsistencies that have not been thought to an end. In fact this “feature” can easily turn into a bug because they have made some over-simplifications that are not documented.
Akin to writing a manual “tank-driving for noobs”.
Scrutinizing those models in pursuit of a deeper understanding, alongside with manuals for hardware icons from the internet, I found a lot of unexplainable behavior, which I believe will not be part of the original design. Yet they claim so.
A question that remains yet to be answered is how they managed to squeeze a feed-back design (which most of the units they have there are…) into a feed-forward architecture, because on such a unit you cannot reach huge attenuation ratios despite what the knob setting suggests by design.
The suitable mantra is “machine learned…”
You can theoretically knit your own design or amend existing designs and then upload them to the community, but, as above, none of that is of any use without meaningful documentation. Indeed, you are well advised not to use any undocumented public preset without extensive testing purely from that standpoint.
You cannot enter help files, the uploading mechanism’s help is of little use and no assistance was received from the forum or even a direct request to customer support.
Program help is nicely done, but rigid and very often ambiguous. Some terms have been chosen unwisely and obviously by somebody, whose native language is not English. Again, no additional information is available and suggestions were ignored. The videos help only to an extent. I have offered assistance but with no response.
This all could be a shining gem if it were not treated so step-motherly. I get the impression that their customer support does not know what to do with me. None of the subjects I brought up are unearthly. The few informations that can be found involve extensive detective work.
Make your own decision.