Circular & Panning Delays with H9’s UltraTap

[marcusm750]

While I doubt there’s any hope of truly pulling these off with the TimeFactor, the H9’s Ultratap algo could possibly be used to realize true circular and panning delays if and only if level and fader controls can be added for each tap.

Towards that end, what about adding a grid feature to the algo to set the gain and panning for each tap?  The Resonator’s grid selects the chomatic note and subdivision; the EQ-Compressor’s grid sets the gain and frequency.  For the UltraTap, the X-axis could represent the stereo spread and the Y-axis could represent the gain.  With a possible 64 taps, this could get ugly quickly so limit the use of the grid to, say, six taps or less to keep it manageable.  Colors and pop-ups could discriminate between the various taps and show the gain/pan values (like in the EQ-Comp).

The grid would conflict with the Taper so that would either have to be set to zero or greyed out.  Likewise, Chop would have to be turned off or greyed out (I don’t really see a practical use for this control in this context anyway).  The Slurm could act as the diffusor and be set to taste as would the Tone (although no high or low filtering is typically used with either of these delays).

For circular delays (three taps), with Spread set to zero for constant spacing, the third tap at 876 ms would be close enough to the typical 888 ms found in the legendary PCM-70 (ie. 292/584/888 ms) that I doubt anyone would notice.  Similarly for panning delays (two taps), the second tap at 752 ms is close enough to the PCM’s typical 786 ms to be negligible for most users.

With the added gain/pan grid, the UltraTap could also be used for a definitive Binson Echorec emulation, complete with all the head permutations and enhanced stereo capability!

[marcusm750]

Really?!  No one would find individual level and pan controls for (some of) the taps useful?

I would think adding the grid in the H9 Control and provisions in the UltraTap algo for the gains/pans wouldn’t be that hard.  I’m not asking this for all 64 taps, just the first six to eight.

We have two PCM 70s at the studio (one the “magic” V2, the other V3) and FilterPong does not sound the same; neither does the UltraTap set up like a circular.  Without the panning, the circular simply don’t have the same width or space; it doesn’t move.  Try it with headphones and the difference will be immediately recognizable.

By the way, still lovin’ the CrushStation! 

 

[Ytolyccuz]

This would be a great addition to the H9! Eventide, hope you’re listening. This could add new textures and ambience in our pallette! 🙂 Looking forward to hear from the Eventide geniuses. 🙂

[skywriter]

marcusm750 wrote:

Colors and pop-ups could discriminate between the various taps and show the gain/pan values (like in the EQ-Comp).

Hi Marcus, nothing agaist your idea per se, but on behalf of all the color blind folks out there – 8% of males – I am 100% agaist the use of color (other than grey-out) to convey any information whatsoever. It’s bad enough that the world of bi-color and tri-color LED’s are inaccessible to us, but you can’t fight cost reductions. However, GUI oriented software functionality can always be implemented without resorting to color. PLEASE DON’T USE COLOR ANYWHERE. Thanks! And now back to our regularly scheduled program!

[marcusm750]

Sorry, SkyWriter, I don’t know of any other way to easily discriminate information on the grid.  Besides, the precedent has already been set in the grids for Resonator and EQ-Comp.  A pop-up can still be used when hovering over the tap to show its number, gain and pan values.

[skywriter]

Hi Marcus, I can’t see what you’re talking about with Resonator and EQ-Comp. Can you please provide a screen grab? thanks!

[skywriter]

Hi Marcus, I checked out the two examples you mentioned. I never used that aspect of the GUI, I did it by ear. But since you brought them up, I see no other reason to introoduce color in these charts other than to help the user in two ways:
1) Resonator – helps the user to associate the setting made within the ‘grid’ to the key. The assistance is provuded by use oof a director line connecting the location to the corresponding key in athe horizontal domain. To me the use of color is redundant with the line, and couple easily be sub stituted with a hashed line with equal benefit.
2) EQ compressor – usues color to help discriminat lines that cross and diverge in a non-ambiguous way. Here, the line color provides a clear assistance that the usualing identical lines themselves cannoct. However, once again using a hashed line to differentiate between converging and diverging lines can provide the same discrimination between such lines as color provides. I can’t tell Gain 1 from Bass, but I can hear the different.

While some may not like hashed lines over colored line they provide the same functionallity to all users regardless of handicap. Also, there are certainly other solutions available to address this software issue; different palettes. With pallete colors; if used should afford the user complete use control, so I can select colors to prevent color aliasing.

The real nature of the two example you selected demonstates something very important in using color in the GUI: the color is only there to help define the information on thre graph better. A benefit here that had been discussed yet is that thee IDENTITY of the color is NOT IMPORTANT. This is crucial, if the user was required to not only discriminate the colors (not a biggie in my book) but to IDENTIFY them as Red, Green, Yellow, Oange, Purple, etc… it would be a complete failure. A little known fact about color blindness (Red/Green in particular), is that being able to name the color is often magnitudes more difficult than to merely differentiate them.

Here’s a little factoid, that maybe argued with, but certainly demonstrates the problem.
1) A person wiith R/F blindness see roughly 2/3 of the color you see.
2) Of the remaining colors 1/2 will alias with the previous 1/3 missing colors.
3) While, mathematically you might say at least we can see, and agree on 1/3 of the total number of colors. The sad fact is that those remaining colors literally don’t occur in isolation in nature – they will contain, to some degree some percentage of all those other colors we disagree with, can’t see, or perceive in some other fashion.

I have 12 MoogerFooger pedals, they use those tri/bi-color LED’s, and the all look the same color. While the pedals sounds great, it’s on of the worst UI designs with LED’s I’ve ever seen. This is sofware, there are so many ways to do it right for everyone.

Anyway, Marcus no hard feelings, and thanks for gving me a platform to express myself for Eventide.

[st.bede]

My thought would be to create a whole new algorithm dedicated to such types of delays. That algorithm could include an underlator type of mod and thus creating some killer movement across a stereo field. The question would be, how to create an algorithm that is also different and useful for people running mono set ups. (If Eventide keeps on this path with killer H9 algorithms, then Eventide will become the stompbox to have. Sculpt has really up the bar).

[kimonostereo]

I’d love to see circular and pan delays on the H9. It’s one of the reasons I bought it in the first place. Had I known this wasn’t really possible, I may have just held off purchasing it. Still waiting and hoping though!

[mjahoger]

I would love to see this feature too! Hopefully by replying to this thread, it will be noticed by the Eventide programmers. Maybe it is covered by the new algorithm that is being hinted at? Fingers crossed…

[mjahoger]

I wonder whether the Eventide team has seen this thread and are going to release a Circular Delay algorithm to the H9 any time soon?

[mjahoger]

I was able to recreate a circular delay using two H9s with the first H9 set up as a mono delay of 888ms with some feedback, and the second H9 as a panning delay set for 584ms on delay A and 292ms on delay B with neither delay A or B having any feedback. I am still experimenting which (combination of) TFactor algorithms are best suited for the circular delay sound. Hope this helps some of you.

[marcusm750]

mjahoger wrote:

I was able to recreate a circular delay using two H9s with the first H9 set up as a mono delay of 888ms with some feedback, and the second H9 as a panning delay set for 584ms on delay A and 292ms on delay B with neither delay A or B having any feedback. I am still experimenting which (combination of) TFactor algorithms are best suited for the circular delay sound. Hope this helps some of you.

Yeah, this is what Steve Lukather does these days (since ditching his rack gear) with three Hardwire DL8s, one for the panning and two in series for the circular.

The circular is a good approximation.  I run my TimeFactor mono into my H9, stereo out into a pair of Fender Princeton Reverbs spaced about 10 feet apart.  Since there’s no diffuser, I use the Vintage Delay in both the TF and H9 with heavy bit crush and top-end roll off.  With the amps spread out, it really does achieve a swirling effect which, after all, is the whole point.

A slightly more accurate method is to run the three delay lines in parallel to simulate the multi-tap.  This requires a mixer on the back end (I use the good ol’ Rane SM26), however, you get better control of the panning and can also implement the -3 dB attenuation on the 292 ms (right) tap.

[mjahoger]

marcusm750 wrote:

mjahoger wrote:

I was able to recreate a circular delay using two H9s with the first H9 set up as a mono delay of 888ms with some feedback, and the second H9 as a panning delay set for 584ms on delay A and 292ms on delay B with neither delay A or B having any feedback. I am still experimenting which (combination of) TFactor algorithms are best suited for the circular delay sound. Hope this helps some of you.

Yeah, this is what Steve Lukather does these days (since ditching his rack gear) with three Hardwire DL8s, one for the panning and two in series for the circular.

The circular is a good approximation.  I run my TimeFactor mono into my H9, stereo out into a pair of Fender Princeton Reverbs spaced about 10 feet apart.  Since there’s no diffuser, I use the Vintage Delay in both the TF and H9 with heavy bit crush and top-end roll off.  With the amps spread out, it really does achieve a swirling effect which, after all, is the whole point.

A slightly more accurate method is to run the three delay lines in parallel to simulate the multi-tap.  This requires a mixer on the back end (I use the good ol’ Rane SM26), however, you get better control of the panning and can also implement the -3 dB attenuation on the 292 ms (right) tap.

Over the weekend, I also was able to get some good sounding panning delays using the Resonator algorithm by setting the delay time to the maximum value, the rythm of the taps to 2,4,6,6, and the resonance at 0. Some reverb can be added to have a little bit of “smear” on the delays, and the filters for each delay line set close to zero, to keep them dark sounding.

Unfortunately, it is still not the same as having a true multitap/plexstyle delay…

[mjahoger]

As the Univibe requests are now covered by the new Harmadillo algorithm, will the next H9 algorithm be a [4 or 6] multi-tap delay with control over the level, lengt, feedback, panning, and modulation for each tap? This way, users will not only be also to recreate Holdsworth-esque multi-voice choruses, but als to dial in many of the multi tape-head delay recreations that have been released recently (i.e., Echorec / Volante, EP3 / El Capistan).

[mjahoger]

Thanks for the terrific TriceraChorus update. It has really blown my, and others’ minds!

I know that you can’t disclose any information on what algorithms are in the works, but can I please bring this request to the attention of the Eventide team again? A 6 or 8-tap delay, or plex-delay is not available on the H9 and would make the H9 even more awesome, as such delays are -at the moment- completely missing in pedals after the Yamaha UD Stomp went out of production.

[Author]

Posts