Spiff and non linearity

Discussion in 'Software Reviews and Tutorials' started by pratyahara, Jan 1, 2021.

  1. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    Watching closely to Spiff graphs in Plugin Doctor one can see that they are tweaking their graphs (making them much more even than they really are).
    When you make EQ 'linear' on their graph, there is still +.5 dB of gain at 10 kHz (the old trick, so that it 'sounds better')
    Their fixed zero gain is in fact at +2.5 dB (the old trick, so that it 'sounds better').
    With controls given you can never make it completely amplitude linear. The best you can do looks ridiculous on their graph [see attachment].
    So it seems like Oeksound uses cheap tricks, like almost everybody else on the market.
    Spiff Closest to Linear.png Spiff Defaults.png Spiff his linear.png
     
    • Interesting Interesting x 9
    • Disagree Disagree x 1
    • Funny Funny x 1
    • List
  2.  
  3. Friendelek

    Friendelek Producer

    Joined:
    Nov 28, 2020
    Messages:
    167
    Likes Received:
    121
    You have to consider signal type generated in Plugindoctor. Maybe that's why spiff start to process that frequency. Need to check it.
     
  4. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    Well, here's a test with delta of pink noise input. It clearly shows the same EQ 'preferences' as shown with Plugin Doctor test. Obvious even with their default EQ set to +1.5 dB @2kHz to 'bridge the gap'.
    Spiff Pink noise delta.png
     
    Last edited: Jan 1, 2021
    • Interesting Interesting x 3
    • Disagree Disagree x 1
    • List
  5. Friendelek

    Friendelek Producer

    Joined:
    Nov 28, 2020
    Messages:
    167
    Likes Received:
    121
    im gonna read manual and I hope to find some info about how it works
     
  6. db100

    db100 Kapellmeister

    Joined:
    Jan 16, 2016
    Messages:
    125
    Likes Received:
    61
    Nice find! What about soothe?
    Maybe check softube weiss plugins and chandler too when they are released. I sense that they use those tricks too like almost everybody else and still the masses fall for it.
     
    • Like Like x 1
    • Agree Agree x 1
    • Interesting Interesting x 1
    • List
  7. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    There are no telltale manuals.
     
  8. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    I have only a few Weiss plugins in archives, and do not use them.
    But the champion in using tricks so far to my knowledge is Slate, very impure, who use several tricks jointly and pass secret dirty processing to 'please the ear'. Their distortion figures are a disaster.
    Among them is also Kush, who produces mishmash plugins as a rule.
    It is not that they are inapt or poor developers, unable to make quality products, they are just cunning.
     
    • Interesting Interesting x 2
    • List
  9. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    One can add Soundtherapy Gullfoss to the list who greatly manipulate with interaction of controls and introduce an omnipresent 'bump' at 16 kHz whatever you chose.
     
    • Interesting Interesting x 4
    • Agree Agree x 1
    • List
  10. db100

    db100 Kapellmeister

    Joined:
    Jan 16, 2016
    Messages:
    125
    Likes Received:
    61
    One should
    One should make a list of developers who use those tricks and which excessively. Afair plugin alliance also does this as they do recycle code a lot.
    And then people wonder about "digital harshness" when these plugins add up all those high frequencys and distortion..
     
    • Interesting Interesting x 4
    • List
  11. LeakedNoise

    LeakedNoise Newbie

    Joined:
    Mar 31, 2020
    Messages:
    4
    Likes Received:
    0
    Could you elaborate a little on that topic? I find some of the Kush plugins, for example Novatron to be amazing. One of my go to compressors to be honest.
     
  12. Maxie

    Maxie Noisemaker

    Joined:
    Jun 14, 2019
    Messages:
    1
    Likes Received:
    3
    There’s no EQ in Spiff. It’s not an equalizer, it’s frequency dependent transient shaper. By adjusting its bands, you’re basically adjusting transient sensitivity for desired band. You should read product manual before posting nonsense.
     
    • Interesting Interesting x 4
    • Like Like x 3
    • Agree Agree x 1
    • Winner Winner x 1
    • List
  13. Adam Ford

    Adam Ford Producer

    Joined:
    Mar 2, 2017
    Messages:
    159
    Likes Received:
    108
    What you say it's correct but not entirely.
    Compressors and transient designers are different even when comparing the same devices because they have non linear responses and wanted or not they change slightly or in a noticeable way the tonal response of a signal.

    Just an example, I love to use 2A comps on material that needs more presence and body because even if it's a compressor, the topology of the circuit usually makes the incoming signal more present and full so... Just my 2 cents.

    I don't know if developers do it on purpose to make their plugins appeal like a soundgoodizer or the emulated machines (in analog emulated plugins) behave that way but it's always useful to know your tools inside out to be able with time and experience to choose what the situation needs.
     
    • Winner Winner x 1
    • Interesting Interesting x 1
    • List
  14. Anubhav Ukil

    Anubhav Ukil Producer

    Joined:
    Nov 22, 2019
    Messages:
    300
    Likes Received:
    143
    Location:
    World
    Well, what you claim to be an EQ is not an EQ. It is a pre-filter or Sidechain EQ.
    So, the whole post is non-sense it seems.

    True.

    Now, @pratyahara For spectral Dynamic Compressions (like Gulfoss or Soothe), check the plugin's activity with steady state signal.
     
    • Interesting Interesting x 1
    • List
  15. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    As shown in the graph, Spiff processes a transient-wise relatively very even signal (like pink noise in my example) as if the pink noise has a 'detectable', rather pronounced transient gain curve set into two groups: @400kHz and @10kHz.
    ----------------------------------------
    There is no dynamic 'band detection' of a large field of transients in Spiff, (i. e. dynamically finding out where the grouping of transients is 'likely'), Spiff detects just single transients, one by one, of course. User has to tell it by using mouse on nods provided where to look and how much gain to apply. So 'adjusting transient sensitivity' is envelope shaping. That envelope regulates gain change, which can be obtained either through EQ-ing or by compressing.
    There is no hyper-intelligent transient sensitivity of Spiff, but the user himself modifies the transient gain regions of his choice.
    The dynamic tests, further, show no compression or expansion, and frequency response tests point at EQ-ing. So EQ-ing it is.
    The problem is that Spiff's envelope can not be made flat by user, and that default envelope state is tweaked by providing added gain in two regions mentioned above - so people who test in on default settings get impressed at the first sight by its 'efficacy'.
     
    • Like Like x 1
    • Interesting Interesting x 1
    • List
  16. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    Now, let us see the frequency response graphs of pink noise before and after Spiff.
    After Spiff there are some minor, hardly visible new transient peaks, but the whole region from approx. 12 kHz to 20 kHz is raised by approx. +1 dB. Transients alone can not contribute that much to the overall frequency response, of course. No Spiff.png With Spiff.png
     
    • Interesting Interesting x 1
    • List
  17. jarredou

    jarredou Producer

    Joined:
    Jan 25, 2017
    Messages:
    154
    Likes Received:
    93
    Interesting post, but I don't understand how can you see transients in a static frequency graph... Peaks in a frequency graph are not transients, they are resonances. Or am I missing something about your tests ?
     
  18. No Avenger

    No Avenger Moderator Staff Member

    Joined:
    Jul 19, 2017
    Messages:
    8,912
    Likes Received:
    6,110
    Location:
    Europe
    I too find it very interesting and even suspicious what you discovered about Spiff, but of course a transient designer can affect a Pink Noise's peak, see

    [​IMG]
     
    • Interesting Interesting x 2
    • List
  19. pratyahara

    pratyahara Rock Star

    Joined:
    Jun 23, 2020
    Messages:
    890
    Likes Received:
    437
    Location:
    Србија
    Yes, but I referred to static cumulative graph.
    Resonance peaks are a bit sharper.
     
    • Interesting Interesting x 1
    • List
Loading...
Similar Threads - Spiff linearity Forum Date
FS | OEKSOUND Soothe,SPiff. Sonnox Dynamic Eq, Envolution (Invoices Included) Selling / Buying Dec 20, 2021
Acustica Diamond Transient vs Oeksound Spiff Mixing and Mastering Oct 22, 2021
Selling Oeksound Soothe+Spiff [sold] Selling / Buying Jan 2, 2021
Loading...