Please ensure Javascript is enabled for purposes of website accessibility Jump to content


  • Posts

  • Joined

  • Last visited

  • Days Won


Posts posted by zolko60

  1. 12 minutes ago, codamedia said:

    Nobody is saying that! 

    I am saying that.

    Before I post anything on ideascale which is read by Helix devs every morning I would like some delay fans/Helix users to confirm my observations.
    Can you also hear signal degradation on Simple Delay or my ears are fooling me?
    Have you played with any vintage digital delay or remember how it crushes repetitions?

  2. 38 minutes ago, lou-kash said:

    Could be a bug. Report it in the bug thread.


    Even if it is. It works like that, nobody complains so why fix it? You can always say it was intended.
    I would just add a few new delay blocks. Names: Dumb Delay, Dub Delay and Brutal Vintage Digital Delay. ;)

  3. 8 minutes ago, DunedinDragon said:


    So why not post this to Ideascale since that's the only place Line 6 developers will ever see it?  They surely won't see it here....


    I have posted it here and on facebook group. The reason is maybe I am wrong at some point and my observations are false?
    The idea of making dumb delays that do not mimic anything or take a look at existing ones supposed to mimic old digital equipment isn't fresh, is it?

  4. Let's talk about Simple Delay vs Vintage Delay.
    There is nothing simpler than Simple Delay, right? Actually no. Line 6 developer has decided that its repetitions are somewhat "modeled" or "bit crushed" and little bit low pass filtered. The feedback goes to 100% which means at that value it gives infinite repetitions. Good but what I hear it does not feed back the crushed/filtered signal. Repetitions remain constant forever. The fun with early digital delays like eg. Boss DD2 was the signal was fed back to poor 12 bit AD converter at very low sampling rate and therefore it was degrading with each repetition. This is supposed to be modeled with Vintage Digital so why alter the sound in Simple Delay? BTW, while checking Vintage Delay I switched to 8 bits and the same behavior takes place. The repetitions are not fed back to 8 bit "crusher". They are crushed once and remain constant forever. (EDIT: I have been listening to it for 20min. It sounds like noiseshapping is used?) Why?
    So what about Dumb Delay? The delay that does not alter the sound at all or alters it if desired with selectable simple EQ curves?
    The last question to Line 6 devs is: why not let the users to use feedback above 100% to make pedal controlled "dub" oscilations?

  5. If you are direct monitoring the recording/playback path is exactly the same so it can't be any difference. However there is a chance you are doing something wrong eg. monitoring direct and software monitoring at the same time. In that case the latency makes comb filter with the direct signal which you may like better.

    • Upvote 1
  6. The problem may be, the avialiable for download Windows driver is not the latest one. As far sa I remember the latest 2.8X Windows driver was only included with the whole update package.

  7. Nice chat on the topic in closed fb group
    Line6 employee claims nothing latency wise has changed with 2.9X firmware. Some people report RTL about 6-8ms on OSX at buffers sizes 64 and 32 samples.
    If anybody can confirm 32 (OSX) and 64 smpl buffers to be usefull (artifact free) please share your RTL reports or measurements and CPU model.

    Thank you for your cooperation!

  8. New idea! Let's help Line6 2.0 - Helix wiki manual
    The Helix manuals (especially FL , LT) are very outdated. We are told writing a new manual is very substantial investment of time and resources. This is probably written by one author Eric Klein (aka DigitalIgloo) and need to be translated to many languages. This is why that work can be done at major releases firmware updates (X.00).
    Useful knowledge as specifications, new DSP blocks decription and their DSP usage, new funcionalities, MIDI manual, that should be in the owners manual, can be found in various places.
    Why not help Line6 and cooperate on the Helix wiki manual when all that content can be published, verified by Line6 employees and then edited and PDF published as neeeded?

  9. 45 minutes ago, Simon268 said:

    What to you think about my approach: To add a fixed gain block directly before the output block.


    It should conserve the author intended level relations between the snapshots ... or am i missing something?

    Yes. You are playing for yourself and for your audience so there is no need to show any respect for the author's level intensions.

  10. 1 hour ago, tahiche said:

    And this is with Helix volume knob at full and no padding? (that's unity I believe). Is this independent of the guitar? 

    The volume knob does not matter with leveling patches. Its purpose is setting global monitoring volume. The unity is also irrelevant to leveling patches.
    The guitar can be a factor. The brigde PAF style humbucker gives about -30LUFS unpadded. How the output loudness is dependent on the guitar is not researched yet. Still some measurements and setting some reference patches solves the problem.


  11. 17 minutes ago, tahiche said:

    How do you keep - 20LUFS at output without a level indicator?.

    If you measure once or believe me that 01A, 01C, 03D are about -20LUFS no matter what you play then you can make a similar volume on any patch.


    21 minutes ago, tahiche said:

    Voicelive 3 has some basic rudementary leds to indicate the level

    ...and Helixes released so far have no leds to indicate the levels.

  12. I see the second topic of this thread is the idea of implementing metering in Helix hardware.
    I can imagine such a need but please consider:
    - we do not know what is the screen refresh or if a display GPU (if any) can handle such a task,
    - we do not know if and what DSP resources would be required,
    also Helix Floor has 7 analog and 12 digital inputs, 8 analog and 10 digital outputs - that is a plenty of meters!

    26 minutes ago, lungho said:

    But as far as clipping goes, even the X3 Live had a clipping light. 

    OK, but if you keep -20LUFS at output, your headroom is quite safe.

  13. There is no unity gain in case of amp modeling. After the amp peak level is likely reduced and rms level is rised. You put about -30LUFS (integrated) signal from a guitar and get -20LUFS at the output. Sometimes you need to be +/-5LUFS louder/quieter in the song context and loudness is heavy playing style dependent, so thinking of the absolute loudness equality hardly makes sense.
    Hx Native can mimic Helix hardware when input and output are set the same. Guitar Input +11dBu full scale (+17dBu with pad) and +11dBu full scale instrument level (+19dBu line level).
    -20LUFS output target is safe, it gives you a reasonable headroom. You can learn to target such a loudness level using any type of metering your DAW or mixer has or even  your ears and compare it eg. to the factory patches of the same type.
    The volume knob is just a digital attenuator used to set the monitoring volume. The main reason of using it set clockwise at building patches is to prevent clipping, but if you keep -20LUFS output loudness level whatever method you develop or choose, it is unlikely to happen.
    There is a thread people present various methods of dealing with their loudness: 


  14. 7 minutes ago, codamedia said:

    I have my method of balancing patches (you can read my posts in this thread if you want) and they work 100% for me

    I thought if you have your method of balancing patches and it works in 100% cases, you can demonstrate it on those three patches and somebody can check if it works for him. For the remote checking it is important to provide dB value of the correction.
    Saving the output block volume value is of course more convinient way of leveling patches than using the volume knob with no scale. This is definitely the idea worth spreading.

  15. 4 minutes ago, codamedia said:

    But there is a simple way.

    The output block has a level control. Balance the two patches and you will never have to touch the volume knob between presets again.

    Balance to what reference on what playing? If the difference is 8 LUFS on loudness metering you can expect even more difference on RMS or peak metering.


    10 minutes ago, codamedia said:

    Problem is... you will soon realize the perceptional differences still exist and there is no automated solution for perception.  

    So if the patches of the different kind can be leveled by perception (ears) please propose the output block volume corrections of those three patches I tested.

  16. 5 hours ago, Parapentep70 said:

    It is impossible to keep the balance in loudness among very different  presets  for different input signals.


    Simple example: If you have 2 simple patches, one with a drastic compressor and NOTHING in the other, if you balance loudness for WEAK signals, when you play stronger the non-compressed patch will sound proportionally louder. And the compressed patch will compress... and sound LESS louder.

    I am just a sunday scientist... I believe the loudness measurement algorithm is immune even to drastic compression/limiting but the guitar distortion is apparently in some pathological ballpark. :D
    My conclusion is strong but if you play "Chugga chuggah" you probably do not play "House of the Raising Sun" on the same patch, so LUFS measurement may be still usefull for comparing the same kind of patches or/and the same kind of playing. This is why I think absolute leveling to equal loudness is not necessary/possible at all.

    My question is still valid: Can anybody prove that leveling (to equal loudness) those three patches makes sense and can be done by ear better regardless different playing?

  17. I have done some research on leveling patches by loudness metering using Youlean Loudness Meter 2 set to read EBU R128 integrated loudness value.


    I have recorded three pieces of guitar playing (about 30s, DI track and then reamped tracks):
    House Of The Rising Sun (HORS)
    Smoke On The Water (SOTW)
    D drop Chugga Chuggah (DDRP)


    I have used three 2.81 firmware factory patches:
    01A US Double Norm
    01C Brit Plexi Brt
    03D Revv Gen Red 


    Absolute Measurements:


    Title       DI     01A   01C   03D   (LUFS)

    HORS  -30.7 -17.4 -19.0 -21.0
    SOTW -31.9 -23.3 -20.2 -21.9
    DDRP  -33.9 -24.7 -20.3 -20.3


    Loudness differences with 01A as the reference:

    Title    01A  01C  03D   (LUFS)

    HORS  0.0  -1.6  -3.6
    SOTW 0.0  +3.1  +1.4
    DDRP  0.0  +4.4  +4.4


    As you may notice on "House of the Rising Sun" patch "Revv" is 3.6 LUFS quieter and 4.4 LUFS louder on "D Drop" with respect to the patch "US Double Norm" with different playing. 8dB (LUFS) difference is the serious issue.



    Using the loudness metering (EBU R128 integrated) for leveling very different patches with very different input material may not be the accurate method or require some additional "weightning".

    How do you judge the loudness of those three patches with your ears?

    • Upvote 1
  • Create New...