Source engine recompile errors

I decompiled a model from Gmod (that was ported to SFM), but when I tried to recompile it, I got errors. They told me that the model had too many facial flexes for source engine (I tried using studiomdl for HL2, L4D2, CSS, CSGO, and all of the source SDKs. Since I wasn’t too concerned about facial flexes, I removed all of its info in the .QC file. But it gave me an error saying the model used too many bones for source engine. I tried compiling with all the studiomdl’s again and nothing worked. So, My questions are: Is there a way to bypass these limits at all? Does anyone have any idea how the model may have even been compiled for gmod in the first place? Like I said, the model was originally made for SFM, but was ported to gmod, so is there a special process to doing this that may have lead to these issues?

If needed, this is the model: http://steamcommunity.com/sharedfiles/filedetails/?id=593261147
The workshop uploaded did not make the port, he only uploaded it. I asked him if he compiled it and he said he didn’t. He told me a guy named “detreter” from deviantart ported it, but when I asked him about a month ago, he never responded.
Thanks in advance for any info.

I heard somewhere that using the DMX format (instead of the SMD format) allows for more facial flexes at the cost of compatibility with older branches of the engine. Confirmation needed though.

I see. I’ll get to converting real quick and see how it goes.

[editline]23rd April 2016[/editline]

When I compiled with .dmx over .smd, I still got the flex error “Too many flex controllers, max 96”
However when I removed the flex data from the .QC, it compiled fine, so the bone limit problem is solved. However it could just be because I don’t really know proper syntax of .dmx. Do I just use the original .vta file and all of the original text in the QC?

Hello. I am the guy who ported it. If u messaged me on DA, i have not read my messages in a long time to avoid being distracted (and most of them kill my brain cells).

This port took about 40 minutes one afternoon to do. So it was a sort of rush job. Now I am no expert in the intricacies of DMX and SMD, but there are certainly compatibility issues going backwards, since DMX supersedes SMD. DMX is simply easier for me to use, as a porter.

Now, what i am **guessing **is leading to these issues is, decompiling. Decompiling DMX’s to SMD’s seems to lead to messed up SMDs (only flex-wise, naturally. thats where DMX differs). But who knows. Indeed i do know the issues you are talking about, i have encountered them as well. The only way I know to avoid them is to avoid decompiling DMXs and if you want a model, try to obtain the original source files.

I think of DMXs as a one-way street. If you want to make adjustments, download the original SFM version from aardvark and go from there.

EDIT: you did not specify what 3d program you use. I know the blender workaround, since thats what i use. If you are trying to do this by editing text files, it will be a world of pain

Don’t even need to ask me, since I happened across this thread.

Click here to download the XNALara Quiet port’s source files

If you want the source files for the other Quiet I released on SFMLab, just ask me over at my Tumblr, or message me through SFMLab.

I’m always more than willing to share my source files. All you need do is ask.

Alright, many thanks guys. Really helped me out.
And yes, I am using blender, however I’m not entirely sure how everything works, all I know how to work is the import and export button