So just wondering why combo or powerstack output modes on hd500 via l6 link to dt50/25 is the thing to do and turning off cab modelling
Just to clarify: combo/poweramp and stack/poweramp do not turn of the cabinet modelling; they only turn off the microphone modelling. And yes, those modes when listened to via direct recording, or through a full range / flat response PA speaker or similar monitor will sound 'fizzy', 'buzzy', etc. - especially when using stomps, drive, distortion, etc.
And yeah, the studio/direct mode is the ideal mode for sending direct signal from the HD500 to a mixer or monitor, but it's not ideal for use as your sound going to an amplifier, and that boominess / low end boost can be overwhelming.
Try simplifying your approach somewhat, in terms of what you are doing.
Set up an HD500 patch with no FX, just an amp model - maybe start with the Plexi.
Set up your gear, use the L6Link, etc.
Now, imagine you are plugged into a Plexi, and dial in the gain, eq, and volume. Get a tone that sounds good to your ears; it should sound good; at this point, shouldn't sound fizzy or weird or anything, and if it does, something else is wrong.
When I first got my HD500, as an upgrade to the X3Live, I literally just played for hours, auditioning only the amp models, before I even tried a single FX block. Even now, I create very minimalist patches, with as few FX possible, since the purity of the amp tones is so amazing.
Anyhow, from there, put a '63 spring reverb into the post section, put a dynamic delay after it. I usually make those FS3 and FS4.
Put a compressor up front, boost comp or tube comp, I make that FS1, and put a distortion after the compressor, I make that FS2. Put a wah first in the signal chain, make sure it's set to EXP1/2 for control and toggle/on/off.
Audition each effect one at a time, tweak the levels and settings - most of the default settings are good, but they will all require some amount of adjustment. When you have each FX block sounding good, levels balanced, on or off, no radical volume or tone shifts, try turning on all the FX blocks. Make sure doing that doesn't squash your sound or triple your volume.
Save that patch. Then save-as into the other patch blocks in that bank. Go to the newly saved copies of your patch (or do all of this with HD edit) and for each one, select a different amp model. Try the tweed, the blackface, the Marshall Park, the SLO - whatever you like.
Repeat that first part, where you adjust the amp settings so each patch is approximately volume leveled, and adjust so that you can select patch A,B,C,D - each one with all FX off, but with those four FX assigned as FS1,2,3,4.
When you are happier with the sound of each amp model by itself, then mess with the FX settings so that each amp model sounds good with those same four FX blocks.
Whenever possible, try to imagine the amp you are using in the model as a 'real' amp somewhere. The DT should sound like that amp. The issue with hearing studio/direct is you are adding multiple variables, and each amp model has different settings, primarily the mic models. Some of the mic defaults sound louder / quieter than others, some will be boomy, some lots of treble, etc. A condensor mic will behave differently than an SM57. A dynamic mic will sound different than a ribbon mic.
The reason you go into combo or stack/poweramp mode, is to disengage the mic modelling. That mode allows your HD and DT to focus on replicating that particular amp model and all it's correspsonding nuance, without also simulating the room and microphone.
Imagine it like this: In Studio/Direct mode, your amp is in another room, with a mic on it, and you are in the control room, behind glass walls, and the mixer monitor speakers are what you are hearing. Essentially a recording / front of house mix sound. Your amp + A Microphone, in a room.
In combo/stack poweramp mode, you are in the room with your amp, and it's next to you. In this mode, the HD and the DT are transforming your DT into a Fender, or a Vox, or a Plexi, etc. The XLR out from the DT is now handling the idea of 'mic' and 'room', whereas what is coming out of your actual speaker is the sound of an amp. The direct signal from the XLR out of the HD500 in this mode will not sound good, the XLR out from the DT will sound good -it's designed to emulate the cab+room+mic.
The reason I say 'imagine' this, is by visualizing the difference between a small 1x12 Gibson combo, and a 4x12 Mesa Boogie is massive, in actual volume, and EQ. You want to pretend you are plugged into the amp that the HD+DT is trying to emulate.
For what it's worth, try an amp model you like, do the copy to four patch idea - then instead of changing the amp model, just change the cabinet for each of those four patches. Again, somehow the poweramp modes emulate cabinet response, while not emulating microphone response - so that EQ which you tweaked to perfection on one cabinet choice, will sound radically different on another cabinet choice.
I've found that the key for me to get a tone I can play music with, rather than just play with the patch editing, is to narrow down my choices to just a few amps and effects, and manipulate patches with these limited selections. Once you are happy / familiar with how those amps and all their options respond, then you can delve deeper into what the HD+DT has to offer. Just randomly switching from amp model to amp model, you will find drastic volume differences, which when taken in the "real world" context of the actual amps which Line6 used to create these models, makes perfect sense.
Put a VOX 25watt 1x12 next to a 100watt 4x12 JCM, and yeah, one is going to be way louder than the other!
Last, but never least: Have Fun!