Waves V-Series versus UAD Neve plugins

[Plec]

Just as a side note...

If you want to see some harmonic generation, then load up the SSL Channel Strip instead of the EQ... that will start to give you some harmonic generation. But that's due to the dynamics in the chain and not thanks to the EQ or Analogue button. With the same 220Hz 0dBFS sine you start to get some harmonics peeping through at 660Hz, -60db, and that's perfectly normal for a compressor starting to work.

This is only for Waves SSL of course, have no idea about the V-Series, but I'd be surprised if it was any different. :lol:

 

[Paul Woodlock]

Well.. sorry to tell you this, but you guys seem to be misled here. I can understand the need to defend your gear, but in 3 different null-tests I've never had anything left when comparing analogue button on/off, other than the low-level noise it leaves behind.

Just for fun I did a quick fourth... using a 220Hz sinewave at 0db (which shows any added harmonics quite easily)

*Sine wave vs. the same phase reversed = Complete Null
*Sine wave with SSL EQ vs. Unprocessed phase reversed = Complete Null
*Sine wave with SSL EQ and "Analogue" button engaged vs. Unprocessed phase reversed = Low-level noise without any signal visually or audibly related to the source material.
*Sine wave with SSL EQ and "Analogue" button engaged +11db and countered with -11db at the master fader vs. Unprocessed phase reversed = Low-level noise without any signal visually or audibly related to the source material.

If Pigcat for example got a 60-70% cancellation with music, I should've got A LOT of harmonics left on this sine-wave thing, especially at that level since harmonic generation varies with level and 0dbFS sine-wave is waaaaaaay hotter than ANY musical material you would ever hear, and you know pushing a 0dbFS sine wave through a clean preamp would inject a considerable amount of harmonics into the signal since optimum operating level is at about -14dBFS. The noise of the Waves SSL stuff is down at -120 to -130db or so... any added harmonics would show up instantly over that noisefloor. AND take into consideration that the SSL 4000 is a very "dirty" sounding board for being solid-state, which a model should generate harmonics being anything but subtle. :lol:

Basically, people need to learn how to make correct null-tests.
...

Plec is correct. You dson't even need to do null tests. Just use a spectrum analyser like IXL. Noise at around -125dB and that's it really

 

[Pigcat]

First of all I need to clarify the result of the null test I did. I did mean the ratio of the leftovers after the null test, after normalize, yes as usual 0dBFS noise, but there's *constant* music of 30~40% of the overall level. It's not like appear then disappear, but the whole music is there. Noticed that I put +38dB gain at the master buss, that'll avoid the too-low digital resolution due to too-low levels. Mind you, I'm using Cubase for this test so there's ADC, moreover the G-EQ has 0 latency so basically any DAW should perform the nullt test in the correct way.

And as I mentioned, V-Series did more obvious thing than the SSL. You can clearly hear that there was added harmonics when the analog button is engaged, and instantly disappeared when it's off. I can even listen the effect with my crappy M-Audio card with decent headphone.

 

[klaatu]

Well.. sorry to tell you this, but you guys seem to be misled here. I can understand the need to defend your gear, but in 3 different null-tests I've never had anything left when comparing analogue button on/off, other than the low-level noise it leaves behind.

Just for fun I did a quick fourth... using a 220Hz sinewave at 0db (which shows any added harmonics quite easily)

*Sine wave vs. the same phase reversed = Complete Null
*Sine wave with SSL EQ vs. Unprocessed phase reversed = Complete Null
*Sine wave with SSL EQ and "Analogue" button engaged vs. Unprocessed phase reversed = Low-level noise without any signal visually or audibly related to the source material.
*Sine wave with SSL EQ and "Analogue" button engaged +11db and countered with -11db at the master fader vs. Unprocessed phase reversed = Low-level noise without any signal visually or audibly related to the source material.

If Pigcat for example got a 60-70% cancellation with music, I should've got A LOT of harmonics left on this sine-wave thing, especially at that level since harmonic generation varies with level and 0dbFS sine-wave is waaaaaaay hotter than ANY musical material you would ever hear, and you know pushing a 0dbFS sine wave through a clean preamp would inject a considerable amount of harmonics into the signal since optimum operating level is at about -14dBFS. The noise of the Waves SSL stuff is down at -120 to -130db or so... any added harmonics would show up instantly over that noisefloor. AND take into consideration that the SSL 4000 is a very "dirty" sounding board for being solid-state, which a model should generate harmonics being anything but subtle. :lol:

Basically, people need to learn how to make correct null-tests.

This is the result of the last test-run. If you can find a trace of a 220Hz 0dBFS source signal or any harmonic related to it... I'd be more than happy to admit that I'm wrong. (Watch your speakers! 0DBFS NOISE!!!)
http://www.panicroom-productions.com/eq ... d_Null.wav

Very nice written. Read this folks and dont be fooled. I own SSL bundle and it is GREAT. However someone in waves think that people dont know what they are doing

 

[lux]

Im must be tripping because when I put the SSL bus comrpessor on the Master buss and turn the analogue button on and off I can hear a difference in the sound.With the analogue button on the sound is more smeared and with it off it is clearer more precise.

Also Ive noticed when driving the input hard on the SSL E strip and bring down the output the sound seems more saturated.

 

[Plec]

First of all I need to clarify the result of the null test I did. I did mean the ratio of the leftovers after the null test, after normalize, yes as usual 0dBFS noise, but there's *constant* music of 30~40% of the overall level. It's not like appear then disappear, but the whole music is there. Noticed that I put +38dB gain at the master buss, that'll avoid the too-low digital resolution due to too-low levels. Mind you, I'm using Cubase for this test so there's ADC, moreover the G-EQ has 0 latency so basically any DAW should perform the nullt test in the correct way.

And as I mentioned, V-Series did more obvious thing than the SSL. You can clearly hear that there was added harmonics when the analog button is engaged, and instantly disappeared when it's off. I can even listen the effect with my crappy M-Audio card with decent headphone.

Thx for the clarfication.

Keep in mind that the result of this null has been boosted by 120db and not a mere 38db. Still can't hear a trace of the source signal. If any useable musical harmonics would've been created on this it would have been at least 70-60db louder than the noise floor. Considering the actual level of the source material, harmonics at 100-110db louder than the noisefloor (10-20db softer than source material) would've been expected.

Could you please explain what you mean by ADC here? If you mean Analog to Digital Conversion then I got to ask what kind of null tests you are trying to do. I hope you're not expecting to null a signal that has gone through an ADC with a signal that hasn't, even trying to null two signals that has gone through the same converters won't do it either. Latency is then an issue no matter what, not to mention the coloring of the ADC itself. :lol:

lux:
Try do a null test to see if you actually get anything left. What you end up is always the difference between two versions of audio. You easily fool the mind into believing what you expect it to, even though there might not be any physical difference. It's purely psychological... period. I also though that the analogue button added "something" when I first tried it, so I did a null-test just to clarify what I was hearing. BIG SURPRISE!

If you hear that difference on the SSL E, it has a logical explanation. That is since the dynamics processor is always in the chain, it adds harmonics to the signal even if the ratio is set at 1. So driving the input harder DOES generate more harmonics but has nothing to do with the analogue switch.

 

[Pigcat]

Could you please explain what you mean by ADC here? If you mean Analog to Digital Conversion then I got to ask what kind of null tests you are trying to do. I hope you're not expecting to null a signal that has gone through an ADC with a signal that hasn't, even trying to null two signals that has gone through the same converters won't do it either. Latency is then an issue no matter what, not to mention the coloring of the ADC itself.

Hope the 0 latency I mentioned right after it gives you some hints, it means automatic delay compensation, or known as PDC which I should be using for 'easier' understanding. By the way, this shouldn't affect the test anyway, as both channel has the same insert, with just 1 of them have the analog button switched off (although it can be 1 with insert and one without). However, I feel like someone in UAD think that people dont know what they are doing, too. :roll:

As for your test, sorry that doesn't prove the fact. It does not take a single test to prove things NOT happening, but only take 1 test to prove it IS happening. Moreover, these are just plugins, not analog gear, so don't compare them with analog gear, simple as that. They might not work with all materials, but it does add harmonics on some.

Whether the analog button is useful or not, that's a different story (which I hope you don't mean it within the topic as discussed, but judging from how you laught over statements in your posts...well... :roll: ), I would say it's utterly useless apart from giving smearing and noises which can help on bringing digital sources to life but that's rare occasions and that might be more useful on high samplerates. Most stuffs are recorded with analog gear thus introducing noises.

 

[Plec]

Whatever makes you feel more comfortable. :lol:

 

[Macc]

both channel has the same insert, with just 1 of them have the analog button switched off (although it can be 1 with insert and one without).

Err, if there's no plugin, how can you do the test you described?

I did the test, insert EQ over a stereo mix, both gain +12dB and fader -12dB, duplicate the channel with the analog button off, phase invert, master buss gain +38dB, and finally normalize the file.

....

However, I feel like someone in UAD think that people dont know what they are doing, too. :roll:

If the cap fits...

these are just plugins, not analog gear, so don't compare them with analog gear, simple as that.

Putting a big red button with 'ANALOG' on it makes that comparison unavoidable, really, don't you think?

 

[Pigcat]

Err, if there's no plugin, how can you do the test you described?

It's the same with no insert and EQ with every knob set flat + analog off.

....

What's your problem?

If the cap fits...

:roll:

Putting a big red button with 'ANALOG' on it makes that comparison unavoidable, really, don't you think?

No. We're in a discussion of what the analog button do.

Ah well. UAD forum is probably the worse place to discuss Waves. Way too much off topic shits.

 

[Macc]

My mistake - I misunderstood which control you were putting at +12, I am with you now. I thought you meant the eq controls all at +12.

No. We're in a discussion of what the analog button do.

You are the one who said 'don't compare them', I merely pointed out that Waves are inviting that comparison with a button like this.

Ah well. UAD forum is probably the worse place to discuss Waves. Way too much off topics.

People here have been pretty fair I think - doing tests and so on. A lot of UAD customers are also Waves customers and would like to know exactly what the situation is. It just happens to be being discussed on the UAD forum, big deal.

Plec's test and Paul Woodlock's analysis seem fairly incontrovertible, but I'll do some of my own tests and make my own conclusions.

 

[Pigcat]

Thanks for the clarification Macc.

You are the one who said 'don't compare them', I merely pointed out that Waves are inviting that comparison with a button like this.

My point is, what we're discussing here is to find out if the analog button is doing what Waves claimed. So to compare it with hardware is literally off topic here. UAD plugins did claimed that they emulate the non-linearities as well, apparently it doesn't to the gain driving tricks with harmonics especially those tube gears, no?

You're right with what will people do when you see that big red Analog button there, but Waves doesn't claim the analog emulation behaves exactly the same as the hardware where different gain/leveling took part. It'll literally blow everyone off the water if it really does the trick where you can mess with input with the hardware. But at least it does something that are accurate of what they claimed. At least, in my test, it did. However I wished they'll release an update with the analog button doing all those non-linearities where you can drive the gain to get the same reaction like how the hardware does - be it "faking" or faithful reproduction. I doubt that'll ever happen, but it'll be interesting for sure.

Anyhow, I really just wish to see opinions from anyone who at least demo'ed the UAD Neves and Waves V-Series, so let's just stay on topic. By the way I'll get a chance to try out the UAD Neves by myself one day...

 

[Macc]

I wished they'll release an update with the analog button doing all those non-linearities where you can drive the gain to get the same reaction like how the hardware does - be it "faking" or faithful reproduction. I doubt that'll ever happen, but it'll be interesting for sure.

That would indeed be sweeeeet

By the way I'll get a chance to try out the UAD Neves by myself one day...

Are you buying a UAD card then?

 

[lux]

First of all I need to clarify the result of the null test I did. I did mean the ratio of the leftovers after the null test, after normalize, yes as usual 0dBFS noise, but there's *constant* music of 30~40% of the overall level. It's not like appear then disappear, but the whole music is there. Noticed that I put +38dB gain at the master buss, that'll avoid the too-low digital resolution due to too-low levels. Mind you, I'm using Cubase for this test so there's ADC, moreover the G-EQ has 0 latency so basically any DAW should perform the nullt test in the correct way.

And as I mentioned, V-Series did more obvious thing than the SSL. You can clearly hear that there was added harmonics when the analog button is engaged, and instantly disappeared when it's off. I can even listen the effect with my crappy M-Audio card with decent headphone.

Thx for the clarfication.

Keep in mind that the result of this null has been boosted by 120db and not a mere 38db. Still can't hear a trace of the source signal. If any useable musical harmonics would've been created on this it would have been at least 70-60db louder than the noise floor. Considering the actual level of the source material, harmonics at 100-110db louder than the noisefloor (10-20db softer than source material) would've been expected.

Could you please explain what you mean by ADC here? If you mean Analog to Digital Conversion then I got to ask what kind of null tests you are trying to do. I hope you're not expecting to null a signal that has gone through an ADC with a signal that hasn't, even trying to null two signals that has gone through the same converters won't do it either. Latency is then an issue no matter what, not to mention the coloring of the ADC itself. :lol:

lux:
Try do a null test to see if you actually get anything left. What you end up is always the difference between two versions of audio. You easily fool the mind into believing what you expect it to, even though there might not be any physical difference. It's purely psychological... period. I also though that the analogue button added "something" when I first tried it, so I did a null-test just to clarify what I was hearing. BIG SURPRISE!

If you hear that difference on the SSL E, it has a logical explanation. That is since the dynamics processor is always in the chain, it adds harmonics to the signal even if the ratio is set at 1. So driving the input harder DOES generate more harmonics but has nothing to do with the analogue switch.

Im not into doing null tests Im happy with my ears even if I am tripping out

 

[fishtank]

Before Mr. Woodlock and everyone else congratulates Plec for showing how the Waves Analogue button does nothing but add noise (and thus proving how awful Waves products are etc. etc.) we need to look a little more closely at all of this and not just jump to conclusions. Plec has demonstrated the tendency to make assumptions that he shouldn't be making. I myself have made an assumption recently that I should not have made - that all the Waves SSL plugs had added distortion by using the Analogue button.....it appears the SSL EQ does not add distortion just as Plec stated - BUT!!!! The SSL Comp sure as hell does!!! First of all, I cannot even recall ever using the SSL EQ - just the Channel and Buss Comp. I knew damn well I could hear a difference when I engaged the Analog button on the compressor and it was not just noise - it was definitely added distortion. After reading Plec's post I knew something had to be wrong as I know what I heard. Surprising to me, the SSL EQ did indeed only seem to add noise and not any distortion, but when I performed the test with the SSL Buss Comp I did see the distortion I was expecting (and you can see this distortion even with the compressor bypassed by disengaging \"IN\" button on the plug but leaving the \"ANALOG\" button on). I used two Group channels in Nuendo with an SSL plug in inserted on each while inverting the polarity on one. I drove them both identically with a pair* of test tone generators (using single and two tones for testing). By turning on the Analog buttons o and I off I could easily see the distorition - the same distortion that would cancel if BOTH Analog buttons were engaged. I tried this with the Ren Comp as well and you could easily see the distortion that the \"Warm\" button adds on that plug.

*One more thing....you may want to use two or more test generators on different freqs for these test as well. I could see IMD showing up with the Analog and Warm buttons engaged. Music is not just a single pure sine wave and the added IMD (in small doses) will thicken things up a bit.

 

[Pigcat]

Are you buying a UAD card then?

Nope, I have a few friends own the UAD card, but I don't know if they own the Neve or still have the unactivated demo. I always wanted to buy UAD plugins, but the latency definitely a no go for me as I use PT M-Powered to compose along with some VSTi and my keyboard. I've however heard some pultec and 1176LN and they're awesome!

it appears the SSL EQ does not add distortion just as Plec stated

I tried this with the Ren Comp as well and you could easily see the distortion that the "Warm" button adds on that plug.

I try to avoid using the compressor for the test for that reason, and I did use the G-EQ for the full spectrum stereo mix test. The leftovers was not like just part of the spectrum, it was the whole music is there - well not really the lower frequencies, you can definitely hear the mid to higher spectrums. Unfortunately it was too subtle to write home about - the music was even lower than the noise so you can imagine how low the level was. Anyhow they should stacked up and become more obvious over bunch of tracks that uses them. Of course, noise add up too and become more obvoius before the music did... :?

 

[svs95]

I have both, and I like them both. Actually, if I had to say which one adds more character, I'd choose the V-series. If you ask me which is a more faithful emulation, I'd say the UAD.

The V-series seems a bit rounder and more diffuse. The UAD version, a bit more precise and clean.

If I'm in a hurry, and the source material is really good to begin with, and therefore isnt' going to be \"fussy,\" I find it a little easier and faster to dial in a really good sound with the V-series. There is definitely a kind of \"vintage glow\" about the sound.

Some material benefits from a cleaner sound like the UAD Neves, but don't take that to mean I find them \"sterile.\" Just a little less \"sepia-toned\" and nostalgic, and a little more like a really well-maintained, later production model Neve.

The UAD-1 Neve 1073 and 1084 are still a world away from the Cambridge in terms of warmth and character, while the V-series goes even further into the \"romantic' range.

Hope that helps.


svs95

 

[Pigcat]

The V-series seems a bit rounder and more diffuse. The UAD version, a bit more precise and clean.

That's basically what I thought too, but only judging from the video of UAD 1073, but that said a video like that won't really do a good reference. I did messed around with the audio file extracted from the video, I kinda messed with the similar settings of the 12khz high shelf, and found out the UAD 1073 seems to give a more clean boost and it seems more of the elements hidden inside has been brought out, whereas the V-Series seems to be more roundish and kinda saturated, or smearing, dare I say - not in a bad way though.

Love the mid of the V-Series especially VEQ3, it really gives that vibey analog feel on most vocal tracks I tried, especially on 1khz~3khz range.

 

[svs95]

Well, even though the source audio you mentioned might be rather poor, the way it responds to processing should still give you the general idea.

I forgot to say that I would be hard pressed to pick a \"favorite,\" or \"the best.\" I actually like having both since it gives me options I would otherwise not have.


Stephen

 

[Paul Woodlock]

Before Mr. Woodlock and everyone else congratulates Plec for showing how the Waves Analogue button does nothing but add noise (and thus proving how awful Waves products are etc. etc.)

Well I think the Waves SSL plugs sound great, so that urinates on your little dig doesn't it

we need to look a little more closely at all of this and not just jump to conclusions. Plec has demonstrated the tendency to make assumptions that he shouldn't be making. I myself have made an assumption recently that I should not have made - that all the Waves SSL plugs had added distortion by using the Analogue button.....it appears the SSL EQ does not add distortion just as Plec stated - BUT!!!!

The SSL Comp sure as hell does!!!

Sorry but it sure does not.

First of all, I cannot even recall ever using the SSL EQ - just the Channel and Buss Comp. I knew damn well I could hear a difference when I engaged the Analog button on the compressor and it was not just noise - it was definitely added distortion. After reading Plec's post I knew something had to be wrong as I know what I heard. Surprising to me, the SSL EQ did indeed only seem to add noise and not any distortion, but when I performed the test with the SSL Buss Comp I did see the distortion I was expecting (and you can see this distortion even with the compressor bypassed by disengaging "IN" button on the plug but leaving the "ANALOG" button on). I used two Group channels in Nuendo with an SSL plug in inserted on each while inverting the polarity on one. I drove them both identically with a pair* of test tone generators (using single and two tones for testing). By turning on the Analog buttons o and I off I could easily see the distorition - the same distortion that would cancel if BOTH Analog buttons were engaged.

1] Test one - Signal Generator into Waves SSSL bus comp, monitoring visually on IXL spectrum analyser.

Result. Switching ANALOG button in only adds noise

2] Test Two. Null test with music - Two identical channels with music track. One channel reverse polarity. - Start with both Bus Comps ANALOG disabled. There's a slight residual amount of the music track when monitorign volume turned way up. Enable ANALOG button on one channel and absolutely no change. I coudln't hear the added noise at -135dB becuase it's below the residual noise level of my montioring system.

Result. Switching ANALOG button in only adds noise

-----------
I then tried the same two tests wth the SSL Channel strip.... Same result... Switching ANALOG button in only adds noise