Diva vs Analogue - a real world test

[IncarnateX]

epilogue:

arf arf

And will the owner kindly get this dog out of my backyard?

And please do it fast before I am tempted to kick it over the fence.

Thanks.

Huh?!

Joke Dog or not, you are definitely the one with the most emphatic responses here. And that is stated without any irony

[fluffy_little_something]

Nah, some of the differences were obviously due to sound design, not to synth capabilities

Please demonstrate which differences were obviously due to sound design

For instance the second sound in video 2.
I think there is more square in the Oberheim, to be heard especially towards the end of each take.

In the first sound in that video one can hear that the voice distribution on Diva varies, with the Oberheim it is alternating between l and r with both takes, with Diva it is l-r-l-r-l and then l-r-r-l-l in the two takes respectively. I only hear the difference on my headphones, though.

With patch three there is stronger detuning in the Oberheim sound.

Sound 4 sounds more spacey and finer on the Oberheim.

Sound 5 seems to have a different envelope setting, the Oberheim seems longer somehow, less linear.

With Sound 7 there seems to be a lower cutoff on the Oberheim, so the sound is more muffled.

With sound 9 it is the other way round.

Strange, now that I have listened really close on my headphones, I prefer B most of the time But that has probably to do with patch settings and not with the synths as such.

[SWAN808]

I asked this simple question on another forum, but never got an answer iirc.

I did answer the question - maybe you didnt read it...you are reaching for ways to criticise the test and this is another tenuous one.

The sounds I used were simple bread and butter sounds. So you suggest the OB8 cannot handle these bread and butter sounds and the test is biased towards Diva? So at this point in order to criticise the test we are having to suggest not to use typical analogue style sounds because Diva is better at those...ok....

.............

As for the idea that the test was one big advert - again - Ive been testing stuff for years - and I already did big tests with no commercial aspect. I recommend anyone try to make many very close comparisons with an analogue synth and see how long it takes. The pack idea came about way into the experiment as a way to try to compensate all this time. As anyone can see - this is my first pack. If I was a known sound designer I might understand the skepticism - but actually Im a known tester and analogue enthusiast (on GS at least under SWAN808)...

Having made the test - of course it is stupid not to maximise the project and make it good with positive promotion. But that does not impact the integrity of the test. The only small bias I have spoken about on GS - in that when I made a selection of comparative sounds - I picked the closer examples. This is sensible IMO because Diva and the OB8 are different synths - so you cannot do an exact parameter or component test. That should be obvious in a Diva vs OB8 comparison. Despite being an analogue enthusiast I still feel it is right to give Diva the best possible chance to prove itself against an analogue synth - rather than setting it poorly making differences obvious. The criticism that I made Diva sound 'better' or 'too good' is actually a compliment to me and also Diva-IE it can sound that close or even better. If you think about it there is no 'sound better parameter' so what the result is is that simply Diva sounds better on a very similar patch.

But anyway - even if you try to make Diva sound 'bad' (try to reveal perceived flaws) its still quite close - and still people fail the tests...(even experts). I know this because Ive tested it last year. When I have a theory - I actually test it. I didnt publish the result because it was failed by some well known Gearslutz and it was a private test.

And finally as Ive said many times - if you wanted to do a scientific test the way to do it is to take the wavs, put them in the linked ABX software and do at least 10 rounds per sound (9/10=high confidence result). Of course no one did that and most people on forums can't be bothered to take a test in that way - hence A or B and videos. Forums are best for discussion IMO. Not for finally solving issues in a scientific way. But the option was there.

[Touch The Universe]

lol, now people are switching what sounds best to whatever the hardware samples were psychology is a funny thing no

[Mutant]

And is an additional proof the the sounds were different. ( If they weren't, why would people prefer this or that ? )

Nice to see an intelligent person here.
The farther it is from 50/50, in either wrong or right direction, the more different the sounds are.
But they still were very close, i would say that they were undistinguishable in the mix and with effects.
So OP: good job !

[Lotuzia]

I asked this simple question on another forum, but never got an answer iirc.

I did answer the question - maybe you didnt read it....

No I missed it, sorry for that, and I dont really want to reread the entire other thread. . So the answer is ?

Did you use the soft synth in the first place to compose your *disco track* ? Then just replaced the presets with their OB-8 counterparts

Or

Did you compose the track on the OB-8, then replaced the sounds by the soft synth presets ?

----___

Then for the rest of your post, I just said that presets dont match that well on both machines, and that this *could* explain the differences in perception by listeners, and why some favoured this or that version. And well, that's all.

[fmr]

And is an additional proof the the sounds were different. ( If they weren't, why would people prefer this or that ? )

Nice to see an intelligent person here.
The farther it is from 50/50, in either wrong or right direction, the more different the sounds are.
But they still were very close, i would say that they were undistinguishable in the mix and with effects.
So OP: good job !

The question is, when emulating a specific analog synth, can you still hear differences between the analog synth, and the digital emulation of that specific synth.

I think it's proven that we can... except that, many times, the one we find the best is the emulation

As I said already a few days ago

[SWAN808]

I asked this simple question on another forum, but never got an answer iirc.

I did answer the question - maybe you didnt read it....

No I missed it, sorry for that, and I dont really want to reread the entire other thread. . So the answer is ?

Did you use the soft synth in the first place to compose your *disco track* ? Then just replaced the presets with their OB-8 counterparts

Or

Did you compose the track on the OB-8, then replaced the sounds by the soft synth presets ?

----___

Then for the rest of your post, I just said that presets dont match that well on both machines, and that this *could* explain the differences in perception by listeners, and why some favoured this or that version. And well, that's all.

I made a dotted line after your quite to signal a seperate comment made on the other queries in the thread recently.

The track was made from an idea not jamming on a particular instrument. They are sat next to eachother and the sounds were made on both at the same time. It was a rough emulation of the french house style that is suitable for synth style tracks. As I said in my previous answer - the sounds are typical of analogue synths - IE unison bass, PWM stab, 2 detuned saw Osc arp/topline. What is it about these sounds you think significantly suit one or the other analogue poly synths or Moog Source? The result is simply the way each sounded on those typical sounds.

It is interesting that Diva performs well, people are looking for bias to explain why it was chosen as preferable. I never said it performs best in EVERY case. In fact if you were reading the threads you will see me advising people to listen to the individual sound .wav and spend a bit more time evaluating if they thought A sounded better. This is because as I said - I think putting the Song test first may have biased the overall results - for the reason that I agree with many that Diva in the Song test sounded better. On the other hand I am not so sure with the individual sounds test. But the test is not so much about what I think - its there for others to listen to and discuss.

[zvenx]

...............
It is interesting that Diva performs well, people are looking for bias to explain why it was chosen as preferable. ........

+1
rsp

[Lotuzia]

I asked this simple question on another forum, but never got an answer iirc.

I did answer the question - maybe you didnt read it....

No I missed it, sorry for that, and I dont really want to reread the entire other thread. . So the answer is ?

Did you use the soft synth in the first place to compose your *disco track* ? Then just replaced the presets with their OB-8 counterparts

Or

Did you compose the track on the OB-8, then replaced the sounds by the soft synth presets ?

----___

Then for the rest of your post, I just said that presets dont match that well on both machines, and that this *could* explain the differences in perception by listeners, and why some favoured this or that version. And well, that's all.

I made a dotted line after your quite to signal a seperate comment made on the other queries in the thread recently.

The track was made from an idea not jamming on a particular instrument. They are sat next to eachother and the sounds were made on both at the same time. It was a rough emulation of the french house style that is suitable for synth style tracks. As I said in my previous answer - the sounds are typical of analogue synths - IE unison bass, PWM stab, 2 detuned saw Osc arp/topline. What is it about these sounds you think significantly suit one or the other analogue poly synths or Moog Source? The result is simply the way each sounded on those typical sounds.

It is interesting that Diva performs well, people are looking for bias to explain why it was chosen as preferable. I never said it performs best in EVERY case. In fact if you were reading the threads you will see me advising people to listen to the individual sound .wav and spend a bit more time evaluating if they thought A sounded better. This is because as I said - I think putting the Song test first may have biased the overall results - for the reason that I agree with many that Diva in the Song test sounded better. On the other hand I am not so sure with the individual sounds test. But the test is not so much about what I think - its there for others to listen to and discuss.

Thanks for the answer. So, in short : Presets did not sound identical, but rather similar - like in identical = 100% and similar can be anything from 50% to 99,99. I let people decide on the pourcentage-. You did not succeed in making the OB8 sound *right* ..... for your track. I posted some videos where some people did succeed in that somewhere in this thread.

And yes, you said many times that the soft synth could not replicate the OB8, I agree. ( Sorry for the *best/worse* contest you bring back once again, this time on an individual clip basis, its not really my concern. )

Now, OB8 is not exactly my most desirable old lady tbh. But still, it can sound very nice, has its own chanting & delicate flavor, and I'd love to get one someday. But probably never will. So I'll use other real analog dinosaurs, or soft synths, like you. Possibly some different ones.

With this I'll leave this thread, because I dont have anything else atm to add. I'm happy I guessed correctly. ( And not by random, as you know )

Best luck with your sales.

[Synthetic Wav]

And is an additional proof the the sounds were different. ( If they weren't, why would people prefer this or that ? )

Nice to see an intelligent person here.
The farther it is from 50/50, in either wrong or right direction, the more different the sounds are.
But they still were very close, i would say that they were undistinguishable in the mix and with effects.
So OP: good job !

uhh there clearny not that close ! moniting is probably one of the main issu in test like these.. depending your room ect one or the other might sound better. in a bright sounding room and maybe bass shy i could see people enjoying more A .

my acoustic a little dark, but almost totally flat from low to 3000hz and here B sound clearly better at my place. A sound sort of muffled and muddy in the low mids while B sound better balanced tonaly, punchier and don't need much work in the mix, my room beeing a little too dark, A sound clearly too dark there and with problems in the low mids, it make it sound like it lack clarity/details .

it's one problem i have with the uhe synths, it s kind of love and hate, to me it s not easy to have great sounding highs, maybe cause of that low mids emphasis i don't know but there is something . not that it s bad, it's just a matter of taste, some moog sound dark and people seems to love these , but i also know some who have to use saturator to create hamronics in the highs, equing to get a more even balance on these synth cause it just lack in the top end, like the filter never fully open, it make these synt great for lower octave moslty for me but thats all..) i also don't see the point of dark sounding synth, making a synth bright sounding dark is very easy..but on a dark sounding synth to have great sounding highs is way more complicated , impossible.. better use a synth with flat tonal response to start with , it s more versatile and should need less work in a mix.

[SWAN808]

Ltz I dont think you understand testing if you claim you did not get the answer correct by chance. Your result of a correct answer is a very low confidence result. You cannot claim any authority by it I'm afraid.

https://en.wikipedia.org/wiki/ABX_test#Confidence

I said to you before that I am not saying its impossible to get the answer correct by knowing the sound - I am saying that you cannot claim in any sort of robust way that you did get it right through your own skill/knowledge. In order to do that you need to do the 10 rounds of ABX testing as described - and get 9 out of 10 correct.

Your interpretation of this test is very confused and as Ive said - clearly reaching to diminish it. You say that I was biased towards the software - then suggest I should bias the test towards the OB8 and use patches that are best on the OB8. Few of your responses make logical sense.

[Lotuzia]

Ltz I dont think you understand testing if you claim you did not get the answer correct by chance. Your result of a correct answer is a very low confidence result. You cannot claim any authority by it I'm afraid.

https://en.wikipedia.org/wiki/ABX_test#Confidence

I said to you before that I am not saying its impossible to get the answer correct by knowing the sound - I am saying that you cannot claim in any sort of robust way that you did get it right through your own skill/knowledge. In order to do that you need to do the 10 rounds of ABX testing as described - and get 9 out of 10 correct.

Your interpretation of this test is very confused and as Ive said - clearly reaching to diminish it. You say that I was biased towards the software - then suggest I should bias the test towards the OB8 and use patches that are best on the OB8. Few of your responses make logical sense.

I've never said such things as you described (like you were biased or similar ).

And I had nearly 100% confidence with my -right- vote. Because I have ears. And references. And this can be seen especially on GS. Believe it, or not, I just dont care. ( In case you should think you made a test so good that no one in the world could guess right except by chance, wich shows well your own level of confidence, or self satisfaction, is stratospheric and is far far beyond the 3000% bar )

Now you could think that you're engaged into a fight with me, but this is an error. I'm just the mirror of your possible failures here. Your commercial page is all about you gloryfying yourself with the results of your test, but now the results of the same test would not work just for those who deciphered the truth. Seriously ....

Then, do I claim for authority ? You're the only one here trying to prove something, right ?. Fight alone, my friend, it would not be fair for a star like you to debate with the confused humble me.

[trimph1]

...............
It is interesting that Diva performs well, people are looking for bias to explain why it was chosen as preferable. ........

+1
rsp

Besides Omnisphere this synth has it's critics, haters, jealous people, conspiracy theorists and complainers who, of course, are never biased.

[Aryaroman]

Whoa. What a difference. It's like night & day...



"goes back to Omni..."