Gearslutz thread with negative observations on MSpectralDynamics
-
- KVRAF
- 3645 posts since 3 Nov, 2015
You mean, at that place where people who are not making music meet ?
-
- KVRian
- 896 posts since 2 May, 2015
...mmmm...interesting...there are two parts so far...1/ an observation, quite detailed, about how the plug works, 2/and another with a complaint about an experience with installation and no real details on platform/os/etc...soooooo...1/ i'll try to replicate and see what i get, and 2/ I've installed and uninstall/etc a bunch of times on both windows and osx (both with the installer and manually) without the issues described...no clue...but clearly #2 is not happy...(sigh).../s~
mba m2 15" | 16gig.ram | 1tb ssd | Sonoma 14.2.1 (23C71)
mbp i9 16" | 16gig.ram | 1tb ssd | Sonoma 14.2.1 (23C71)
logic10.8.1 | reaper7.07 | focusrite.2i2
mbp i9 16" | 16gig.ram | 1tb ssd | Sonoma 14.2.1 (23C71)
logic10.8.1 | reaper7.07 | focusrite.2i2
-
- KVRist
- 452 posts since 21 Jul, 2018
Wasn't intended to be a thread with "negative observations." Was intended to be a thread with impartial observations, and questions about how those observations might be rectified with the intended outcome... asked by someone who makes music for a living.
To sum up:
Pushing MSpectralDynamics hard has a rather consistent tendency to tilt the outcome brighter than the controls would indicate.
For instance inputting pink noise into the unit with a 3dB/octave "slope" setting, and a horizontal threshold dropped down to the lowest setting (-80dB) so it's essentially "always on" (zero attack, slowest release (and again, very low threshold) does not output a signal with a 3dB/octave slope.
To clarify: The signal going IN is already 3dB/octave, and the unit is being told to push TOWARD 3dB/octave (which it already is), yet the output is actually FURTHER from the settings than it was on input.
This is not a fluke. I've tested it dozens of times, and always with the same result. For at least 5 tracks now pushed through so that the slope control matches the contour o the infinite averaged peaks of the incoming signal, the output is nearly always something like 1dB/octave BRIGHTER than the input. Always. Happens on music. Happens on test signals. Always. It does appear to be a bit more pronounced in the real world (dance) music examples, though it's always skewed in the same direction.
The only way I've found to counteract this is to essentially ADD 1dB/octave or so to the "slope" setting on the unit, so if the incoming signal is at 4dB/octave, and I want the output to also have a 4dB/octave slope, then I actually need to set the "slope" setting on the unit to 5dB/octave. It's very consistently skewed.
Actually, overall, I spent the better part of a week sending hundreds of test signals through it and measuring the effect of every single parameter in fullscreen visual analysis detail.
Imagine my surprise upon signing up here to be greeted by a poster saying they've never used the plug, yet need to "set me straight."
Please do.
To sum up:
Pushing MSpectralDynamics hard has a rather consistent tendency to tilt the outcome brighter than the controls would indicate.
For instance inputting pink noise into the unit with a 3dB/octave "slope" setting, and a horizontal threshold dropped down to the lowest setting (-80dB) so it's essentially "always on" (zero attack, slowest release (and again, very low threshold) does not output a signal with a 3dB/octave slope.
To clarify: The signal going IN is already 3dB/octave, and the unit is being told to push TOWARD 3dB/octave (which it already is), yet the output is actually FURTHER from the settings than it was on input.
This is not a fluke. I've tested it dozens of times, and always with the same result. For at least 5 tracks now pushed through so that the slope control matches the contour o the infinite averaged peaks of the incoming signal, the output is nearly always something like 1dB/octave BRIGHTER than the input. Always. Happens on music. Happens on test signals. Always. It does appear to be a bit more pronounced in the real world (dance) music examples, though it's always skewed in the same direction.
The only way I've found to counteract this is to essentially ADD 1dB/octave or so to the "slope" setting on the unit, so if the incoming signal is at 4dB/octave, and I want the output to also have a 4dB/octave slope, then I actually need to set the "slope" setting on the unit to 5dB/octave. It's very consistently skewed.
Actually, overall, I spent the better part of a week sending hundreds of test signals through it and measuring the effect of every single parameter in fullscreen visual analysis detail.
Imagine my surprise upon signing up here to be greeted by a poster saying they've never used the plug, yet need to "set me straight."
Please do.
-
- KVRian
- 896 posts since 2 May, 2015
I use MSD fairly frequently and have fell I have a "pretty good" sense of how it works, but I've never done the testing/poking you've described (the OP that is). In my use cases I adjusted the parameters until I had something I liked, however in due time, I think I will try to replicate...it's always good to learn something new...Cheers!.../s~
mba m2 15" | 16gig.ram | 1tb ssd | Sonoma 14.2.1 (23C71)
mbp i9 16" | 16gig.ram | 1tb ssd | Sonoma 14.2.1 (23C71)
logic10.8.1 | reaper7.07 | focusrite.2i2
mbp i9 16" | 16gig.ram | 1tb ssd | Sonoma 14.2.1 (23C71)
logic10.8.1 | reaper7.07 | focusrite.2i2
-
- KVRAF
- 35436 posts since 11 Apr, 2010 from Germany
Dunno about that, but... whenever i get a link to a Gearslutz thread when i Google something, i keep wondering why that place is supposed to be where the "professionals" meet. Maybe i always happen to find the wrong threads, IDK.mevla wrote:You mean, at that place where people who are not making music meet ?
-
- KVRist
- 283 posts since 22 Sep, 2015
Yeah kvr is way better of coursechk071 wrote:Dunno about that, but... whenever i get a link to a Gearslutz thread when i Google something, i keep wondering why that place is supposed to be where the "professionals" meet. Maybe i always happen to find the wrong threads, IDK.mevla wrote:You mean, at that place where people who are not making music meet ?
-
- KVRAF
- 35436 posts since 11 Apr, 2010 from Germany
Should i link you to some horrific GS threads? No, really, i always find the kind which makes you happy that you're on KVR, and the people at least have remotely an idea what they're talking about. Really. Seriously.
-
- KVRist
- 283 posts since 22 Sep, 2015
Okay then I'm obviously looking often into the wrong ones here I always felt like on kvr are more cry babies (having problems that need to be said as often as possible in as many threads) and on gearslutz are more critics (discover problems a bit more scientific but don't want to discuss problems rather than using other tools).
Last edited by Soundplex on Sat Jul 21, 2018 8:35 pm, edited 1 time in total.
-
- KVRAF
- 35436 posts since 11 Apr, 2010 from Germany
Hey, could well be that i also typically find the threads with peeps who have no idea on the Gearslutz forum. As i wrote though, i haven't found one decent topic yet. The ones i visited were all crowded with esoteric nonsense.
-
- KVRAF
- 2596 posts since 9 Jul, 2015 from UK
I will run some tests and get back to you soon. Everyone is really friendly and helpful here.
I think the comment about setting them straight was aimed at the second post, where someone was calling the installation like a virus, which had absolutely no relevance to your question at all.
Anyway, welcome to the forum, glad to see another Melda power user!
I will try and get an answer to this for you.
I think the comment about setting them straight was aimed at the second post, where someone was calling the installation like a virus, which had absolutely no relevance to your question at all.
Anyway, welcome to the forum, glad to see another Melda power user!
I will try and get an answer to this for you.
Melda Production & United Plugins
Surface Studio = i7, 32gb, SSD.
Windows 11. Bitwig, Reaper, Live. MTotal.
Audiofuse, Adam Audio monitors + sub, iLoud MTM.
Polybrute, Summit, Pro 3, Tempest, Syntakt, AH2.
Ableton Push 2, Roli Seaboard Block.
Surface Studio = i7, 32gb, SSD.
Windows 11. Bitwig, Reaper, Live. MTotal.
Audiofuse, Adam Audio monitors + sub, iLoud MTM.
Polybrute, Summit, Pro 3, Tempest, Syntakt, AH2.
Ableton Push 2, Roli Seaboard Block.
-
- KVRist
- 283 posts since 22 Sep, 2015
You mean that discussion that came up here many times and comes up with any new updatejmg8 wrote: think the comment about setting them straight was aimed at the second post, where someone was calling the installation like a virus
-
- KVRist
- 452 posts since 21 Jul, 2018
I had already "set him straight" on the virus issue prior to this being linked.
Thx.
At first, I thought it was something along the lines of the kick/bass being loud, but not constant, so even though they show as equal on an average energy curve, their short intense bursts receive more total GR than the more sustained sounds higher up...
... except that it's not greatly distorting the curve so much as just tilting it, and the same (albeit slightly reduced) tilt shows up on pink noise.
Thx.
At first, I thought it was something along the lines of the kick/bass being loud, but not constant, so even though they show as equal on an average energy curve, their short intense bursts receive more total GR than the more sustained sounds higher up...
... except that it's not greatly distorting the curve so much as just tilting it, and the same (albeit slightly reduced) tilt shows up on pink noise.
- KVRist
- 338 posts since 19 Jul, 2013 from Chile
Well, as a rather long time Melda user, and much longer audio-related various forums participant, I have to state that I have only admiration for Gearslutz forums.
They have been probably the most authoritative ones in many pro audio regards, starting by the dedicated sections where some of the most important music producers of all times have had their time personally responding questions from the users. I personally value Q&A with Bruce Swedien and others. Plus some of the big hardware brands representatives have been there directly responding questions threads about their products, something very seldom seen elsewhere.
The last years have find myself always checking there when I am finding out about gear that I might be interested in.
That said, no need to underscore either the high regards I have about Melda focus on resolution and high quality of its algorithms... together with the valuable feedback and input from the informed users community here, that is mostly received by the developer if found relevant.
Welcome Annabanna.
They have been probably the most authoritative ones in many pro audio regards, starting by the dedicated sections where some of the most important music producers of all times have had their time personally responding questions from the users. I personally value Q&A with Bruce Swedien and others. Plus some of the big hardware brands representatives have been there directly responding questions threads about their products, something very seldom seen elsewhere.
The last years have find myself always checking there when I am finding out about gear that I might be interested in.
That said, no need to underscore either the high regards I have about Melda focus on resolution and high quality of its algorithms... together with the valuable feedback and input from the informed users community here, that is mostly received by the developer if found relevant.
Welcome Annabanna.
Last edited by Nspace on Sat Jul 21, 2018 9:22 pm, edited 1 time in total.
-
- KVRian
- 542 posts since 28 Oct, 2014
Really, this is no surprise to me. When i was using MmultibandDynamics (for 2 or 3 years), i was constantly having to compensate for things that shouldn't be happening. For instance, when drawing a custom curve, set to raise the volume of the quietest areas (less than -20db), the plugin insisted on raising the volume of the the whole band (0db transients included) by a good few decibels.
As far as i could tell, at 0 attack and 0 release, this shouldn't be happening, and when i talked to Vojtech about it, he just insisted that i shouldn't be using the plugin that way.
Although Melda do make my all-time fave plugins, i'm not blind to some of their small quirks.
And seriously guys, jumping in to defend KVR against GS... who cares
As far as i could tell, at 0 attack and 0 release, this shouldn't be happening, and when i talked to Vojtech about it, he just insisted that i shouldn't be using the plugin that way.
Although Melda do make my all-time fave plugins, i'm not blind to some of their small quirks.
And seriously guys, jumping in to defend KVR against GS... who cares