Mailist to discuss all issues regarding the Digitech GSP-2101 Guitar FX Processor

Pedro,

That you're thinking of would be quite an endeaviour to do! You're 
talking about modifying the internal circuitry to source a digital 
signal directly to the processor skipping the A/Ds. Someone else 
please correct me if I'm somehow wrong, but that's anything but 
trivial.

After Saber540's comments about his trial a few days ago, I did my 
own since I also wanted to compare the same thing on my ART multiFX. 
Here is my experimental set-up:

- My "raw" signal flow was as follows:
  Guitar -> Behringer 1600 ProCom compressor -> Digitech 2101LE 
input -> Digitech 2101LE FX loop send -> Mesa Triaxis input
- Compression, Distortion and analog EQ were all switched OFF on the 
2101. Volume control after analog section was set to 0dB on the 
2101. Global EQ was set totally flat on the 2101. 
- Outputs used on the Triaxis were the "record" ones.
- The Triaxis' record left OUT was routed to a Behringer UB1832-
FXPRO mixer mono IN channel directly and panned full left. This was 
my baseline signal.
- The Triaxis' record right OUT went to the 2101's FX loop left 
return.
- All digital FX modules but Noise Gate were removed from the 
algorithm on the 2101. Hence, the 2101 should be acting as just 
a "direct signal pipe" with nothing affecting the signal "in theory".
- Digitech 2101LE left main output went to another Behringer UB1832-
FXPRO mixer mono IN channel and panned full right.
- Main Behringer UB1832-FXPRO mixer outputs went to my SB AWEGold 
PC's soundcard.

I noticed three main things:
1) I can confirm what Saber540 found out: There is a 5ms delay in 
the analog signal coming out of the 2101.
2) The two signals (raw left and 2101 right) WERE NOT identical. 
This was expected as the 2101 uses 18bit 48kHz A/D/A converters. 
However, the overall signal shape was very similar, just 
slightly "attenuated" or compressed. I have to take a closer look to 
the Noise Gate settings to make sure there was no ramp-up effect 
influencing this.
3) Same experience was done with the ART-SGXNitro instead of the 
2101LE. Both signals were totally in-synch, no delay observed, but 
there was a more noticeable difference in the wave shapes. I have no 
info as of what A/D quality the ART uses, so no further conclusions 
can be drawn out of this.

Sonically I must add, while it's true that the raw tube distortion 
from the Triaxis was slightly modified by the 2101 circuitry, I 
would not say that the difference was concerning. Besides that, I 
personally never use a raw signal from the Triaxis, it always gets 
processed with EQ, some chorus, and a bit of reverb or delay, AT 
LEAST.

Are the 18bit A/D converters playing a role on this? I definitely 
believe so, but how much of a difference would you be able to notice 
from a unit like a TC Electronic G-Force with 24bit converters??? 
One more thing: I'm also a believer of using a BBE SonicMaximizer as 
the last component in your signal path. And since there is so much 
difference in the sound when you turn those things on, I'm not 
really concerned about the 2101LE degradating or affecting somehow 
the original signal. The BBE seems to make everything sound great!

I'm so happy having so much flexibility from using the Triaxis as my 
main distortion source (via the FX loop) and the internal 
2101 "Grunge" or tube distoritions whenever I feel like using them 
that the 2101 has been just the perfect solution to my needs. And 
all this is available just at the touch of a button, without any 
other external gadget or anything!!! What else can you ask for??? Oh 
yeah.... I know..... Digitech, could you please launch a 3101 with 
24bit converters???   ;-)

Hope this helps or at least serves as food for thought. Peace!

Gus.

--- In gsp-2101@yahoogroups.com, "Pedro Ferreira" 
<pedro_mig_ferreira@n...> wrote:
> Hello!
> 
> I have bought a high quality pre to go with my 2101, and I have 
noticed that
> the 2101 was not that good, converters related, it takes a lot of 
sound
> quality and attack either I use it though the pre-in or the 
effects-loop-in.
> 
> Does any one know how to upgrade de converters or install a SPID 
or ADAT
> input so that we can inject the signal directly already converted?
> 
> I love this processor, but I might have to replace it if I don't 
improve the
> converters and sell myself to TCelectronic. Please don't let me do 
so.
> 
>  
> 
> Any help would be good.
> 
>  
> 
> Thanks,
> 
> Pedro Ferreira