I used DCP-o-matic to create a DCP package based on the high quality render from my source files (properties=x264-medium f=mp4 vcodec=libx264 acodec=aac g=120 crf=%quality ab=%audiobitrate+'k') but it doesn't seem to quite match the clarity of my original material (from a Cannon C100; ffprobe gives me "Stream #0:0[0x1011]: Video: h264 (High) (HDMV / 0x564D4448), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], 23.98 fps, 23.98 tbr, 90k tbn, 47.95 tbc"). The render to mp4 looked fine on my monitor and looked OK on my 55 inch TV as run through my Bluray player, but when I used the playback function of DCP-o-matic on my PC it seemed to highlight the image softening that I felt like I was seeing on the TV. Like it was slightly out of focus, if that makes any sense.

I found these two links:

viewtopic.php?f=272&t=116177&p=286267&hilit=dcp#p286267
viewtopic.php?f=272&t=117253&p=291047&hilit=dcp#p291047

But they're so old I figured things had to have changed. BTW, I'm using the appimage kdenlive-18.12.1b-x86_64

So are there any current recommendations for creating the best quality output to use for creating a DCP package?

I think it’s more on the DCP-o-matic side then on the Kdenive side. Does DCP-o-matic re-render the MP4 again? Does DCP-o-matic has a “just pass through” function?

Not sure how to answer this. I just run the DCP-o-matic UI, select the Kdenlive rendered mp4 and press go (and wait over 6 hours). I didn't do a side-by-side, but once I noticed the 'softening' in the DCP output I went back to look at the Kdenlive render and noticed the same thing. I believe what brought this to my attention is I've been culling through the original clips looking for outtakes, thus looking at the original imagery, and seeing it's noticeably sharper. I don't know that it would be enough to draw the eye on a TV (my wife didn't notice any lack when we watched it), but my concern is blowing up to a theater screen the fuzziness would draw the eye. I noticed in the rendering options there is lossless, and am wondering if that would retain the original quality, and images sequence. One of the image sequences is JPEG and my research says DCP is basically each frame rendered as JPEG2000, though somehow collected together and synced with sound.

Anyway, just hoping someone out there knows what I'm trying to do better than I do and can point me in some profitable directions.

Thanks!