Circular logic, possible solutions...(long-winded post)

It sounds like you are continually layering the already-recorded and then monitored sound with the sound being recorded. This is because your record-in signal is not being separated from the monitored signal in the mixer. You are trying to have the mixer do two things at once. This can (and will) get confusing, but I'll throw out some ideas to chew on.

There are a couple of things you can try. (I have a Delta 1010. I know the LT is different, but I don't know for SURE that if all of what's next will translate to the LT. Perhaps, entertain the idea, and see if they will).

First, instead of monitoring through your mixer, connect a couple of Delta (left and right) outputs to whatever amp, powered speakers, etc., that you use, and monitor from that. Of course, this eliminates any EQ and level control, etc., that you may be used to using while MONITORING through the mixer....but it also eliminates any noise created in the mixer while monitoring. And, it gives you a better representation of what's on your disk anyway, because you AREN'T tweaking EQs, etc. while you are monitoring the recorded signal.

This also leaves the mixer free to use as an input device, if you so wish. Since the monitored signal is not running through it, it cannot get recorded back in. Of course, you MAY be better off not using the mixer to track at all. Inexpensive mixers may not do your recording much justice, and if you have any means of recording directly into the Delta, it may be a better option. Of course, THAT means you can't use the EQ, effects, etc., of the mixer. But, that may be WHY you want to use the mixer. Of course, monitoring through the computer MAY introduce latency, depending.

Conversely, if you run everything directly into the Delta, bypassing the mixer for recording, then you could use the mixer for monitoring, connecting Delta outputs to the mixer, and the monitoring system (amp, speakers, etc) to the mixer outputs. This means that nothing being RECORDED is running through the mixer until it has gone through the Delta first. And since the mixer is no longer running into the Delta, the signal will be separate. Again, latency may be a factor, depending.

See how this this going 'round in circles?

There are a couple of other things you MIGHT try, though I'm not sure they will work. This next part is what might be possible with certain mixers, and may give you an idea for upgrading the mixer. It doesn't appear there are any "direct out" jacks on that mixer. Sometimes, you can interrupt the signal at that point and direct it somewhere else...like to a recorder...and it won't continue through to the main mix internally of the mixer. Sometimes, you can even specify either post-fader and EQ (uses the EQ and fader of a channel), or pre-fader and EQ (doesn't...signal passes through pretty much unmolested).

That means you could possibly use the mixer to run a signal through to the recording device, and it won't get mixed back into the mixer main unless you direct it back in AFTER it runs through the recording device., if the mixer is capable.
Your mixer has "Auxiliary" inputs and outputs, but the manual says those are post-fader, post-EQ. I think all you can do with that is run something in "parallel", which means the main signal is still going through to the main mix, and the effect is being added to taste, in parallel, with the help of the "Auxiliary" level knob(s). So, I don't think that can be used to separate the signals, (but I may be wrong). A mixer with the capabilities like that would be handy, if you want to use a mixer for tracking and monitoring.

What would happen there is you could plug into the mixer (guitar, mic., etc.) divert the signal FROM the mixer main mix, to the Delta, then monitor the signal along with the rest of the already recorded stuff from the Delta outs through the mixer. Nothing will get looped back through.
The only cavaet with THAT scenario may be that, if you are monitoring everything only through Delta 1 & 2 out...to the mixer...then the signal you are trying to record cannot be controlled separately from the mix, without possibly messing up the recording level balance. In other words, if you want the guitar louder than the mix, and you boost the input at the Delta, you also boost the recording level.

BUT, you may be able to run, say, Delta 1 & 2 outs to the mixer for the already recorded music monitoring, and run Delta 3 & 4 outs also to the mixer to monitor the guitar. ( I THINK you can do this. Correct me if I'm wrong). This way, you can set your levels properly in the computer, and adjust the recorded music/guitar level ratio in the mixer, monitoring without changing levels in the computer. Still, possible latency issues.

What I'm saying is, I don't think that particular mixer has the capability to track and monitor at the same time...not easily, anyway.

You MAY be able to run the Delta 1 & 2 out to the mixer for monitoring the music, and not assign the signal being recorded to a monitor mix out of the Delta...but split the signal you want to record. You would have the signal split to have one path to the Delta input(s), and one path to the mixer input(s).

Play, sing...whatever...and set the levels in the Delta for recording. You could even temporarily assign the recording channels(s) to the main Delta out, and monitor while setting levels, then change the monitoring assignment after setting levels, and before recording, so that it doesn't get monitored through to the mixer. (If doing this, turn down the split signal that is going to the mixer while doing this, or you'll hear both).
Now, you could start playback, adjust your split signal going to the mixer to the level you want to monitor your guitar in relation to the recorded music, and then hit "record".

This has several advantages and disadvantages.
Advantages:
Since the signal is being monitored directly to the mixer, but not through the software, latency won't be an issue. If monitoring a recording track in progress through the computer, as in several of the previous scenarios, latency MAY be an issue. This won't.

About latency, if you aren't having problems with it at already, then all the warnings about latency I have suggested may not be a factor, and you can probably ignore those.

You have independant level and EQ, and effects control of the already recorded signal, and the signal being recorded. Let's say you like to have the "John Lennon slapback" on all your lead vocal tracks, and, like John, find it hard to sing without it. If your signal is split, you run a dry signal with no effects to the computer. BUT, since you are monitoring through a mixer that has effects, (or with a capable non-effects mixer...add effects) you can apply effects to your vocal while you are singing so that you are comfortable, but it won't affect the recorded track. Once an effect is recorded, it cannot be undone.

This has the advantage of letting you sing to effects, and the freedom to apply more appropriate effects to the vocal in the computer at mixing time. You can apply...don't like it...undo and change it. Gives you much more flexability to fit the vocal effects into the mix, instead of just living with what you recorded with effects, that cannot be easily undone.

Disadvantages:
If you don't split a signal properly, you can load down a signal creating noise, weak signals, and other things that cause bad sound. You may also have to worry about grounding issues, etc., because you are running one thing to at least two separate devices, which are connected together. There are ways around that, but you should be aware of it and be prepared to find a solution.

You have to constantly keep changing monitoring assignments in the Delta...depending on how you do it. Now that your track is recorded, you want that track to possibly be mixed into the main Delta 1 & 2 outs for monitoring, for example. Then you have to go through all the level setting stuff with the next track to be recorded, divert it (them) away from the main monitor mix, and constantly remember to turn up or down the split to mixer channel(s) while doing this. Your recording software also will possibly (probably) come into play with assigning and levels, which adds yet another level of possible confusion...but it's in play no matter what scenario you use. You'll possibly redirect the outputs from within the recording software, assuming the Delta Monitor Mixer stuff is set up properly, but the possibility still exists that that you MAY have to open the Delta Control panel. I don't know for sure. (I'm confusing myself, now). Anyway, you don't want to set the levels in the computer that are feeding temporarily out of the Delta into the mixer, while you also can hear the direct to mixer split.
This all can get annoying, fast.

So, some of your choices:

Monitor everything directly out of the Delta, bypassing the mixer for monitoring, and use the mixer for tracking. (Possible latency issues, possible noise added to the recording, possible inferior sound if the mixer is less-than-stellar. BUT, you have mixer EQ and effects, if you wish).

Track everything directly to the Delta, and monitor everything out of the mixer. (Still, possible latency issues, but less noise issues in the recording, but perhaps some in monitoring. No mixer EQ or effects available, but track it well in the first place, and massage the tracks in the software).

Eliminate the mixer altogether, and track directly to, and monitor directly from the Delta. (STILL possible latency issues, but nothing else adding noise into or out of the Delta. Track well, massage later).

Split the signal. (Versatility. Possible noise and other issues, depending. Constant raising and lowering mixer level(s), possibly constant reassignment of Delta outs, etc. Can record dry, monitor with EQ and effects, add effects and final EQ to recorded tracks later in software).

Get a mixer with with more routing capabilities and higher quality components. ($$$$. Can add simultaneous use of mixer for tracking and monitoring. Can also add possibly more confusion and need for patching, or reduce both...depending on mixer).

Patchbay. (May be possible to do a lot of this with a patchbay, but then you would need possibly more cables, adapters, etc. More connections to possibly introduce more failure points or noise induction. Possibly constantly patching and re-patching. Confusion. $$)

There may be some things that I haven't thought of, and I may have gotten wrong a thing or two, but I THINK this is pretty much what you are facing. I just threw some stuff at the wall as they came to mind. Corrections welcome.

Investigate the capabilities of what you have, and perhaps draw up some flow charts of possible routings that would be most useful for you. Perhaps play around with a couple different scenarios to decide what's best.

Hope this helped (more than it confused) :wink:

Kapt.Krunch :shock: