The Legend of Dreamcast Trigger Lag

[undamned]

So, where did this notion of official DC controllers having trigger lag come from?  I remember reading somewhere, when this topic came up, that stuff like Zangief's Lariat wasn't reliable because of this?  Out of curiosity/disbelief I soldered X, Y, and the Right Trigger hall effect sensor pin indicated on the SlagCoin guide together and Zangief's Lariat (MvsC2) worked 100% of the time.

Unless there is gross lag (which I am unable to detect, just playing a game) the only real way to prove lag would be for someone to write a homebrew program for the DC that would sample the button inputs and display response time differences between button presses in milliseconds.  With all punch buttons tied together as I described above, said program could tell us how much extra time it takes for the system to register analog trigger activation over a digital button press.  It can be assumed that the controller's ADC is going to take a bit more time to process the analog sample, but my wager is that it's a delay far too small for a human to detect.
-ud

[NFG]

My first instinct is to blame software calibration:  If you wait for the trigger to be pressed half way, that's a lot of movement and time wasted before anything is picked up by the game. 

Your hypothetical dreamcase software would be a tricky thing to write, since it would have to control the pushing AND record the time taken... 

[undamned]

My first instinct is to blame software calibration:  If you wait for the trigger to be pressed half way, that's a lot of movement and time wasted before anything is picked up by the game.

The context I want to use the DC controller in is in it's hacked form where you are electrically taking the trigger from Zero-to-Pulled instantly, leaving any lag up to the processor.

Your hypothetical dreamcase software would be a tricky thing to write, since it would have to control the pushing AND record the time taken...

It would not control the pushing.  As I said it would only sample the button inputs.   The buttons would be electrically tied together and activated by a human.  The program would be in an endless loop, merely looking for any button press.  It would then record the delay between the first button registered and any buttons which register thereafter (i.e. lag).
-ud

[Endymion]

There are pads out there now that map the triggers to non-analogue, standard buttons, this MadCatz DreamPad for instance, there is absolutely no lag when pressing Y and Z through those buttons. I doubt there is any processor faulted or even controller-borne lag present if you use the triggers, at least not anything moreso than the physical pull of the things which makes them less than ideal from an ergonomic perspective alone for most 2D/fighter type games.

Edit: I'm sorry, that's Z and C buttons, been a while since I looked at mine.

[kendrick]

Not to split hairs, but aren't Y and Z mapped as separate inputs from L and R? I know that the Capcom fighters use both pairs as the fifth and sixth action button, but I think it's theoretically possible to treat them as individual buttons even though no game used that input scheme. 

[undamned]

I doubt there is any processor faulted or even controller-borne lag present if you use the triggers, at least not anything moreso than the physical pull of the things which makes them less than ideal from an ergonomic perspective alone for most 2D/fighter type games.

Technically there will be lag on the analog triggers because it takes the processor's Analog-to-Digital converter more clock cycles to read a transition than just a pure digital input (normal button press).  My argument is that this delay should not be perceivable by humans.  The program I was describing would be merely to solidify exactly how much extra time it takes to process an analog trigger press and prove/disprove whether it is enough of a delay to be perceivable by a human.

Antron, over at shmups, linked me to this thread, which is some sort of a culmination of such rumors (from 2007).
-ud

[NFG]

The buttons would be electrically tied together and activated by a human.

Aahh, so you're pushing two at a time.  I see now.

Technically there may be lag, but since the console and controller only sample at 60Hz, the CPU in the pad is likely running far faster and holding the data until required.  Any internal lag would be irrelevent since the current position is always immediately available to the console.  At least, that's my theory!  =)

[Midori]

Not to split hairs, but aren't Y and Z mapped as separate inputs from L and R? I know that the Capcom fighters use both pairs as the fifth and sixth action button, but I think it's theoretically possible to treat them as individual buttons even though no game used that input scheme. 

Perhaps Virtual On Used that for the Twin Sticks setup?

In a Twin Stick Setup it would require two 4 way directional sticks and two buttons on each stick plus a start button. Those extra two buttons would be useful there if you mapped the first stick to the lines used for the pad and the other stick to the A,B,X,Y buttons. Because then you would only have L,R and the analog stick left for the Four buttons.

[undamned]

Finally somebody pointed me to conclusive testing.  The training mode in CvS2 has a key display option (like in SFIV) which shows you what buttons the console is seeing pushed and in what sequence.  According to Toodles over at SRK, CvS2 registers inputs by video frame, so as much as we don't get accuracy in the millisecond range (as I wanted in my hypothetical homebrew test program), we at least can tell more accurately than your average human could perceive (10's of milliseconds):

http://www.shoryuken.com/showthread.php?t=144153

Looks like Sega never originally intended the L & R triggers for anything other than analog.  Software developers ended up using them for digital actions because they were available and better than being stuck with only 4 face buttons.

That being said, Madcatz was the largest producer of DC pads with 6 pure digital face buttons (C and Z buttons being direct digital mirrors of L & R action).  Be leery of some of the other 3rd party controllers which have a 6 button configuration, but the two extra buttons are merely programmable macro buttons.
-ud