What would you consider a "long enough" delay time for most practical purposes?

[JTEX]

Just curious what others think about this. Is 0.5sec long enough to cover almost all real-life situations where a delay effect is needed? Or maybe 1sec? Or is that already in looper territory (not interested)? Emphasis on real playing situations, real songs, in a band/gig, not just as a party trick or to mindlessly "have fun" with a delay.

 

[Paradox916]

For me I seem to live somewhere between 300-600ms ( depending on tempo) and it’s just my personal preference but anything over 800ms is useless to me personally. I honestly don’t think I ever go above 500ms very much I just tune that in by feel and ear

 

[airbud]

for "normal," 800ms

for "normal ambient," at least 1.2 sec

mostly arbitrary so ymmv

 

[JTEX]

Asking mostly because I can squeeze about 600 to maybe even 700ms out of a PT2399 with reasonable sound quality, depending on filtering aggressiveness. If I had to go higher, I'd have to find some other solution - likely a lot more complex and power-hungry.

 

[bean]

Asking mostly because I can squeeze about 600 to maybe even 700ms out of a PT2399 with reasonable sound quality, depending on filtering aggressiveness. If I had to go higher, I'd have to find some other solution - likely a lot more complex and power-hungry.

The FV-1 will get you 1sec and will sound better than any PT2399 most likely. You can even program it to have PT2399 noise, if'n-u-want.

 

[JTEX]

Yeah, but I don't have the power budget for a FV1. Looking for single digit milliamps. Battery powered.

 

[Paradox916]

Is this just a fun project, or taking it to market?

 

[JTEX]

Both. Everything that I ever put on the market started as a fun project for my own use.

 

[Paradox916]

The 2399 would be fine I think it it was just a project trying something out for giggles, but it seems like people expect more out of a delay than what the pt2399 has to offer, unless you can do something super clever with it. Sorry not trying to be a downer, just my take.

 

[DailyDovetails]

Asking mostly because I can squeeze about 600 to maybe even 700ms out of a PT2399 with reasonable sound quality, depending on filtering aggressiveness. If I had to go higher, I'd have to find some other solution - likely a lot more complex and power-hungry.

For the PT2399 I’m not particularly fond of them on the longer delay times. I’m going to guess maybe 300-400 is around the max for me. That gives you most of the delay times you would want without being set to the tempo but you probably would need longer times for that. 1 4/4 beat at 120BPM is 500ms and at 60 is 1000ms.

 

[JTEX]

I agree that PT2399 is far from ideal, especially when pushed to out of spec delay times. I had to bandwidth-limit it to something like 2 or 2.5khz with a steep rolloff to get clean enough repeats at >400ms or so. I just can't seem to find a better alternative with similar power consumption. I may have to find a really low power DSP with decent built-in ADC, DAC, and enough RAM, and learn how to program the stupid thing. Sounds like a truckload of work to me. I really wish they'd just make a better PT2399 with twice the RAM or even more.

 

[DailyDovetails]

I agree that PT2399 is far from ideal, especially when pushed to out of spec delay times. I had to bandwidth-limit it to something like 2 or 2.5khz with a steep rolloff to get clean enough repeats at >400ms or so. I just can't seem to find a better alternative with similar power consumption. I may have to find a really low power DSP with decent built-in ADC, DAC, and enough RAM, and learn how to program the stupid thing. Sounds like a truckload of work to me. I really wish they'd just make a better PT2399 with twice the RAM or even more.

The old school delays like the DM-2, MM, AD-9 ect all have pretty short delay times. Might be fun to lean into that. Are you thinking of doing modulation?

 

[JTEX]

No modulation. Just a basic delay, as low power as possible while doing >0.5 seconds of delay time at good fidelity.

 

[pingpongcomputer]

Pretty sure the recent Benson delay is PT2399 and sounds awesome plus fairly long delay times, might be worth poking around in one of those

 

[bean]

Yeah, but I don't have the power budget for a FV1. Looking for single digit milliamps. Battery powered.

A single PT2399 delay may put you at 20mA+. That's really not so bad, though even on battery power.

 

[JTEX]

Hehe, I have that one down to under 4mA from a 9V supply, using magic But I wish it sounded more hifi at longer times.

 

[derevaun]

Whatever this is. Actually, it's 300mS. So, 300mS is enough for me.

 

[JTEX]

For reference, this is what my PT2399 delay sounds like at a 600ms delay time. Am I too picky for finding fault with it?

PT2399_600ms

 

[giovanni]

I have found 900ms to 1.2s to be useful in certain situations, mostly for ambient sounds or very slow tempos. But I agree that anything above 700ms is rarely used.

 

[JTEX]

Pretty sure the recent Benson delay is PT2399 and sounds awesome plus fairly long delay times, might be worth poking around in one of those

Surely it must use two PT2399s to achieve those delay times. I'd love to take a look at the schematic, but my Google Fu failed me. They mention using some open source tap tempo circuit in it. One would thing that a product using open source stuff would also be open source. But I guess it depends on the license type.