|
|
Post by craigmorris74 on Oct 5, 2021 21:48:32 GMT -6
I just finished mixing my latest record (mostly with hardware compression), and after reading some recent comments on here about hardware and plugins, decided to do a little test myself to give myself a break. The compressor in question in the 1176 style compressor.
There are two files linked below, at least one of them is hardware. According to many, the answer should be obvious (especially with an 1176)-either both are hardware, or one of them is hardware and one is a plugin. If you listen to the files and tell which is which, I'd love to know your reasoning.
NEW FILES ARE A FEW POSTS DOWN!
|
|
|
|
Post by yotonic on Oct 5, 2021 22:43:44 GMT -6
The two files are not the same volume so that threw me a curve ball. In my experience plug ins always rob some content and dimensionality of the source, similar to a Xerox copy. Your B file sounds more dimensional to me and less collapsed. So I would expect that to be the hardware file. I’m also listening on an iPhone…. Lol. But even on this thing there is an audible difference between the two files.
|
|
|
|
Post by craigmorris74 on Oct 5, 2021 23:36:54 GMT -6
Thanks for your thoughts! Files were level matched with Metric A/B plugin.
|
|
|
|
Post by enlav on Oct 5, 2021 23:44:11 GMT -6
Compression behavior feels different between the two... There's definitely a perceived difference in volume on my monitors.
I'm not sure I know the 1176 enough to even guess (and feel free to shoot me, but depending on the 76-style hardware clone/inspired piece, I sometimes prefer the software over it).
|
|
|
|
Post by Blackdawg on Oct 6, 2021 1:03:06 GMT -6
yeah levels are different, from what I can tell +0.5dB on A matches it damn close.
Pretty big difference actually. Well...not BIG. But noticeable.
I like B better. But Im honestly not confident that's the hardware. But it sounds better between the two..
|
|
|
|
Post by gravesnumber9 on Oct 6, 2021 3:19:57 GMT -6
Snare sounds more open on B... maybe the release is acting faster? I don't know. Either way I like it better regardless of hardware or plugin. The two takes certainly sound pretty noticeably different to my ears.
|
|
|
|
Post by tkaitkai on Oct 6, 2021 9:21:49 GMT -6
I always suck at 1176 shootouts. I feel like that's one area where the good plugins (UAD, Arturia, Slate, Purple MC77) are really close to HW.
In this instance, I agree there's a difference in compression behavior. File B brings up the release more and sounds a bit more slammed than File A, which has harder transients and less decay.
I guess it depends on which specific plugin/HW was used — some plugins can be more squeezed/shaved off like File B, others can be "spiky" like File A, and the same applies to HW.
Anyway, I like B more.
|
|
|
|
Post by craigmorris74 on Oct 6, 2021 13:29:03 GMT -6
Since I am a sick person, I've listened to your comments, and re-recorded the files. I threw in a couple other 1176 style comps for fun. A and B are still the same device.
Also, I used test tones to calibrate gain reduction, attack, and release of each unit. The loudness of each file was calibrated to the nearest 10th of a dB LUFS using the meter of Limter 6 (insane mode for DAN). If you don't think they're loudness matched, they are downloadable, so you can do it to taste.
At least one of the files is hardware, and at least one is hardware. Your thoughts are appreciated!
|
|
|
|
Post by yotonic on Oct 6, 2021 19:53:37 GMT -6
A&C plug ins.
B&D hardware.
B is an 1176 plugged into a wall socket somewhere in western Kentucky based upon the cycles I’m hearing. Clearly the wall socket is running at 116 as versus 120. I suspect that’s because of the beer cooler you have plugged in on the other wall. 😎
|
|
|
|
Post by craigmorris74 on Oct 6, 2021 21:35:24 GMT -6
A&C plug ins. B&D hardware. B is an 1176 plugged into a wall socket somewhere in western Kentucky based upon the cycles I’m hearing. Clearly the wall socket is running at 116 as versus 120. I suspect that’s because of the beer cooler you have plugged in on the other wall. 😎 Very perceptive thoughts, though I can't confirm at this stage the accuracy. I'd like to get the thoughts of the plugin experts, or from those who think plugs make a mix loss it's magic! |
|
|
|
Post by Ward on Oct 7, 2021 5:40:54 GMT -6
A&C plug ins. B&D hardware. B is an 1176 plugged into a wall socket somewhere in western Kentucky based upon the cycles I’m hearing. Clearly the wall socket is running at 116 as versus 120. I suspect that’s because of the beer cooler you have plugged in on the other wall. 😎 Dude, clearly B uses an inferior cheap IEC power cord. You just can't get enough electrons ramping up the volts with those. It's all about the volts, bruh, all about the volts. |
|
|
|
Post by Ward on Oct 7, 2021 5:52:06 GMT -6
Of course I kid . . . but it's subjective because we're not listening to the untreated source, plus it's a mono drum mix. I like 76s on snares and use a variety . . . each one is different, but so is every drummer and every snare. How these elements interact determines the experienced and educated engineer's choice! If I were starting with a Maple 14x6.5 versus a 14x5 supraphonic, I would then determine microphone choices as well as preamp choices. Then start engaging compressors until the drummer's patience wore thin. How a drummer hits determine's the voice for the percussion, and the wrong comp will rob a track of its personality. I played a club last night and the music running before hand was a bunch of recent remakes of songs that were only 10-15 years old to begin with. The snare sounded the same throughout. Just a dull lifeless thud, no snap crackle or pop. No Rice Crispies? I'm an unhappy individual. And also, you want a MUCH quicker release on snare than on kick. Let the kick bloom but allow the snare compressor to recover and get out of the way.  |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 7, 2021 6:05:46 GMT -6
I always suck at 1176 shootouts. I feel like that's one area where the good plugins (UAD, Arturia, Slate, Purple MC77) are really close to HW. In this instance, I agree there's a difference in compression behavior. File B brings up the release more and sounds a bit more slammed than File A, which has harder transients and less decay. I guess it depends on which specific plugin/HW was used — some plugins can be more squeezed/shaved off like File B, others can be "spiky" like File A, and the same applies to HW. Anyway, I like B more. Well plugins have basic issues with non-linear operations, especially compression and even the 1176’s transistor has 100x the bandwidth of 44.1 kHz digital. Most of emulations are hopeless. No way they can get a fast attack coefficient, the splatty dual release, and switching between the attack and release coefficients. The non-oversampled ones like Waves and PSP can’t even see the peaks they’re supposed to be taking off. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 7, 2021 6:07:47 GMT -6
yeah levels are different, from what I can tell +0.5dB on A matches it damn close. Pretty big difference actually. Well...not BIG. But noticeable. I like B better. But Im honestly not confident that's the hardware. But it sounds better between the two.. There’s no right way to volume match. VU, PPU, sample peak, true peak, RMS, LU… etc. especially when the software is not going to be able to take the peaks off and release in the same way. I think the overloud comp 76 goes the highest and that’s not the most popular one. |
|
|
|
Post by Ward on Oct 7, 2021 6:11:18 GMT -6
I always suck at 1176 shootouts. I feel like that's one area where the good plugins (UAD, Arturia, Slate, Purple MC77) are really close to HW. In this instance, I agree there's a difference in compression behavior. File B brings up the release more and sounds a bit more slammed than File A, which has harder transients and less decay. I guess it depends on which specific plugin/HW was used — some plugins can be more squeezed/shaved off like File B, others can be "spiky" like File A, and the same applies to HW. Anyway, I like B more. Well plugins have basic issues with non-linear operations, especially compression and even the 1176’s transistor has 100x the bandwidth of 44.1 kHz digital. Most of emulations are hopeless. No way they can get a fast attack coefficient, the splatty dual release, and switching between the attack and release coefficients. The non-oversampled ones like Waves and PSP can’t even see the peaks they’re supposed to be taking off. Yes. and even after release (when the compressor setting tells the FET to stop working) things happen in recovery (the state at which the compression operator returns to a state of rest). We hear that tail more than we realize, more than we could possibly hear the transient. That's where the tone is. IME |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Oct 7, 2021 6:35:52 GMT -6
Of course I kid . . . but it's subjective because we're not listening to the untreated source, plus it's a mono drum mix. I like 76s on snares and use a variety . . . each one is different, but so is every drummer and every snare. How these elements interact determines the experienced and educated engineer's choice! If I were starting with a Maple 14x6.5 versus a 14x5 supraphonic, I would then determine microphone choices as well as preamp choices. Then start engaging compressors until the drummer's patience wore thin. How a drummer hits determine's the voice for the percussion, and the wrong comp will rob a track of its personality. I played a club last night and the music running before hand was a bunch of recent remakes of songs that were only 10-15 years old to begin with. The snare sounded the same throughout. Just a dull lifeless thud, no snap crackle or pop. No Rice Crispies? I'm an unhappy individual. And also, you want a MUCH quicker release on snare than on kick. Let the kick bloom but allow the snare compressor to recover and get out of the way. Compressors often take the snares out of the snare and they have to be added back in with a high shelf. Modern music doesn’t like high shelves as much due to often having a messed top end to start with Also the “pop” in popular music snares from the last 15 odd years, since at least the Waves CLA Classic Compressors pack for popular music and back in the 90s for electronic, is the compressor not seeing the start of the drum hit on a too fast an attack filter. The 44.1khz samples are not the signal so rectifying them will not rectify the signal, only the samples, and it will produce aliasing in the rectified samples. So will hitting the threshold, a glorified clipper. Then fast filter coefficients will alias horribly and it won’t be able to switch between attack and release correctly. So the signal is being constantly modulated by a side Jackie Chan that’s not working anywhere close to intended and the programmers went with it anyway. Then when dug in, the snares come off the snare, and since the top end is a nasty piece of trash, they can’t be put back on with a high shelf. The exceptions with old digital are the ones that don’t do anything to the high end like Oxford dynamics and waves Renaissance eq which don’t really do anything above 1 and 4 khz except act as a fader rider. Then you can bring back the highs but it’s still not going to attack and release like an analog or modem digital. The software engineers just knew what they were dealing with and got rid of the uh nastiest distortion. |
|
|
|
Post by enlav on Oct 7, 2021 8:15:40 GMT -6
Craig, thanks for all the extra legwork on this shootout/comparison. I'll be honest, with the first two files, I was sort of wondering if it was a trick question and we were comparing the same hardware with just different settings.
On a pair of Senn HD380's, I actually hear some similarities between A and B, and with C and D... which is surprising since I felt a greater difference between the original A and B files, but maybe this all comes from going back and forth through four variations now. C and D both feel like the air/top end is being attenuated in some way, A to a lesser degree, and B the least. I don't trust the 380's enough to really evaluate the low end. (Call me crazy, but I feel like there's more dimension --even though this is in mono-- in B.)
For the context of this track, if this was the end result, no other tampering, I'd choose B; but I'm wondering if I might prefer the others on individual sources/tracks.
My biggest takeaway is how drastic some of these files sound from one another.
A&C plug ins. B&D hardware. B is an 1176 plugged into a wall socket somewhere in western Kentucky based upon the cycles I’m hearing. Clearly the wall socket is running at 116 as versus 120. I suspect that’s because of the beer cooler you have plugged in on the other wall. 😎 Dude, clearly B uses an inferior cheap IEC power cord. You just can't get enough electrons ramping up the volts with those. It's all about the volts, bruh, all about the volts. I don't know, I think if you really focus on the blueberry notes, you could tell this was grown, single-source, on a small high-altitude farm in Guatemala. This crop had particularly higher amounts of rain. The family is splitting up due to jealousy, which I think really comes through as the bitter aftertaste in this brew. Delightful.
Oh wait, sorry, wrong place.
|
|
|
|
Post by Martin John Butler on Oct 7, 2021 8:40:38 GMT -6
B&D. if they're plug-ins, I want one. If they're hardware, what brand, just in case  |
|
|
|
Post by craigmorris74 on Oct 7, 2021 9:10:32 GMT -6
I live right by Kentucky Dam, so the power going through those cheap IEC is pure American hydro, almost straight from the source. I refuse to get juice from the coal-fired plants because we use bituminous coal instead of anthracite in our part of the world (introduces too many impurities for the virgin copper wire).
If you listen closely to one of the files, you can hear a cooling cycle of the beer fridge start.
|
|
|
|
Post by bgrotto on Oct 7, 2021 9:44:21 GMT -6
Since I am a sick person, I've listened to your comments, and re-recorded the files. I threw in a couple other 1176 style comps for fun. A and B are still the same device.
Also, I used test tones to calibrate gain reduction, attack, and release of each unit. The loudness of each file was calibrated to the nearest 10th of a dB LUFS using the meter of Limter 6 (insane mode for DAN). If you don't think they're loudness matched, they are downloadable, so you can do it to taste.
At least one of the files is hardware, and at least one is hardware. Your thoughts are appreciated!
The attack on A seems smooshier (but not necessarily faster) than that of B, and the release a bit slower and awkward in its envelope. That makes A feel kind of sludgy and mushy to my ears. C feels a bit bloated, and similarly mushy to A. D is more open than A and C, and the compression action feels more natural than A and C, but there's something about the low mids that feels a bit wimpier. These differences were fairly apparent coming from my laptop speakers, but i'm keen to listen again in my studio. Subtle tonal differences are not exactly obvious on these tinny little suckers, but the dynamic behavior is still pretty clear. Eager to hear the results! I sorta hope B is a plugin, tbh 🤣 |
|
|
|
Post by craigmorris74 on Oct 7, 2021 9:58:54 GMT -6
A lot of great thoughts posted so far. It was interesting to see how different builds respond differently when set to do the same gain reduction, attack and release.
I want to learn from this thread, though, and I would like someone to point out the digital artifacts we're supposed to hear from plugins:
-smearing the image
-2d sound
-improper handling of the attack
I want to learn to hear this stuff to make myself better.
|
|
|
|
Post by bgrotto on Oct 7, 2021 10:04:55 GMT -6
A lot of great thoughts posted so far. It was interesting to see how different builds respond differently when set to do the same gain reduction, attack and release. I want to learn from this thread, though, and I would like someone to point out the digital artifacts we're supposed to hear from plugins: -smearing the image -2d sound -improper handling of the attack I want to learn to hear this stuff to make myself better. Well...that'll all depend on which of your examples were plugs vs hw 🤣 A and C to me sounds like what i would associate with non oversampled plugin attack behavior (smeary and indistinct). D sounds to me like properly oversampled plugins (a la TDR), but with a more 'digital' tone (hence the low mids comment). I guess that would be called "2d". B sounds like hardware. But I'm very prepared to be wrong on some or all of those predictions...just pointing out what associations I have with what i'm hearing in your test. |
|
|
|
Post by craigmorris74 on Oct 7, 2021 10:14:39 GMT -6
Well...that'll all depend on which of your examples were plugs vs hw 🤣 I'm sure AFTER I lost the results, a lot of folks will identify the obvious digital signature! That's why I really appreciate those who have shared their thoughts so far! |
|
|
|
Post by bgrotto on Oct 7, 2021 10:16:04 GMT -6
Well...that'll all depend on which of your examples were plugs vs hw 🤣 I'm sure AFTER I lost the results, a lot of folks will identify the obvious digital signature! That's why I really appreciate those who have shared their thoughts so far! Ha, yes, they sure will! |
|
|
|
Post by christopher on Oct 7, 2021 10:41:33 GMT -6
I want to learn from this thread, though, and I would like someone to point out the digital artifacts we're supposed to hear from plugins: -smearing the image -2d sound -improper handling of the attack I want to learn to hear this stuff to make myself better. I can’t tell you which is which here. but I can teach you how to hear where analog is required: A: notice how it’s tame and kind of boring? The highs kind of say nothing remarkable, the tone isn’t anything amazing. B: See how hard it grabs and tugs on the sound? It’s so 76 it’s like an exaggerated meme of one. C: Kind of grabby like B, kind of boring like A, but seems sort of transparent in the highs. D: Notice how flat and bland that is? It’s like the whole thing is one generic sound. Each are usable.. but they aren’t interchangeable. A: I want to control a wild source that is already painful to hear, this could get me to tolerable space where I will need to work more. B: I want exaggerated analog grabby feel, source be damned. C: I want the grabby feel but the source is pretty good and I don’t want to destroy it D: I need to glue the thing into one cohesive sound, even though end result will need more work unless I need it to be flat for some reason |
|
