I've had some experience with this, but I'm not a technical artist, so take what I say with a grain of salt. I just go off what "looks" correct to me for use in VRay.
I have given up trying to make the Substance Plugin for 3DS Max part of my workflow for now, because the results are not good for VRay, and it doesn't work with VRay RT. So my method is to simply author my materials in PBR Metal/Rough in Substance Designer, and use the
PBR Conversion Node with the
VRay Target to output bitmaps for VRay Materials (Diffuse, Reflection, Normal, Reflection Glossiness, Fresnel IOR, and others such as Emissive and Opacity if I need them). For me this is desirable, because if I need to send off my work to a render farm, I don't need to worry about whether or not they have the Substance Plugin installed. Honestly, the more plugins you're using, the more likely you're going to run into trouble sending off your projects to be rendered eslewhere, so working with standard bitmaps in 3DSMax can save you a lot of headaches.
There is an issue that has to do with gamma corrections of image formats, when bringing bitmaps into 3DS Max, which has been nicely summarized here already. The image formats I'm referring to have a gamma correction applied to them as they are saved onto your disk (jpeg, png, targa)[Note: If you're using EXRs as your texture format, this does not apply, as they natively have a gamma of 1). I've found that almost all of the maps (normal, roughness [or reflection glossiness if you're using VRay], reflection map, and fresnel ior map, need to have a gamma override (from 2.2 to 1.0) applied to them when you import, or at least need some tweaking in their output curve. This is especially true for the the normal map, which pretty much looks incredibly janky unless your applied the gamma override to 1.0. Strangely, the gamma correction on the basecolor, or diffuse, map, coming out of SD, seems to be fine in most cases. Applying the gamma correction to the basecolor or diffuse map is what is making your material look "washed out," unless I'm mistaken.
There is also the color mapping system in VRay to consider, when it comes to how color "correct," your render is. By default, VRay is setup to use Reinhard, which is a combination of Linear Multiply and Exponential. Now, this can get very technical, and it's still a jumble in my head, but I have generally adopted two approaches, which I'm sure will make some technical artists groan, but I've found that I generally want to do some photoshopping after rendering with VRay 95% of the time, so for me the "correctness," of my render is fairly moot. That is not to say that you shouldn't do everything you can to ensure you get the greatest range of colors in your render, and if the following approaches make no sense for reasons I'm unaware of, I'm all ears. At the end of the day, a photographer is an artist, and a camera is an expressive tool, which can be used to create endless stylistic variations,
I have adopted using the HSV Exponential color mapping type, with the Color Mapping and Gamma mode selected in the advanced rollout. Your render will come out looking washed out or foggy gray, but if you save an EXR(HDR image) out with these settings, I find you have a really great range of values to work with in Photoshop. It surely is a completely inaccurate render, but the results I've gotten from it have been pretty stellar, and require less tweaking in photoshop. I just feel very comfortable with these settings, as I know what I can expect the image to look like after editing.
The second approach is to be as linear as possible, using Linear Multiply, and no gamma correction on the render. This produces a very dark image, but saving out an HDR means that all of the values in your render are there, you just need to bring them out in photoshop, which takes considerable tweaking. You should probably use this method, as it is more correct for sure, but it's just harder to anticipate what you're image will look like before you begin post production. One of the big draw backs of VRay's frame buffer, which kinda made me shy away from the linear method, is that you cannot render LUT's or OCIO color profiles onto your image directly out of the VRay frame buffer. This seems crazy to me, and hopefully a future update changes this, as I'd rather not HAVE to use photoshop, but so it goes.
Hope this was a little helpful
The images I attached are just a quick showcase of how I used the PBR Conversion Node method with VRay Target selected I described, with the Linear Render method. The leather and steel materials was from Substance Source, which I simply brought into a SD graph, and used the PBR Converter on it to get the VRay bitmaps I needed for my VRay Material, then simply rendered in VRay RT
Pretty painless.
