Miguel,
I went through this dilemma recently on a job. We shot 50 mbit MPEG-2 4:2:2 on a Canon C300 (1080 @ 23.976 fps) and transcoded to DNX 80 with the thought of NOT doing an uprez to anything higher. I did this on a show last year with material shot on a Canon 7D and everything worked out fine. No QC kickbacks. But I have thought, with the C300 if going higher to 175 would make any noticeable difference.
I ran this test by re linking to AMA files and re transcoding, then did a comparison in Photoshop between 175x, 80, and 36.
I won’t say there is no difference between 80 and 175x. But it is only perceptible to the eye with 4x or 8x blow up of the image. There is some finer pixel-to-pixel detail that can be seen from that vantage. Mainly in noisier low-light areas. But, overall, at 1-to-1, I could not tell the difference. I would encourage you to run your own test similar to this because it probably varies depending on the source material. It also depends on who you are delivering to I suppose. If your client is stringent on quality, maybe hedge your bets by doing the up res. But if it’s not going to be a big deal, the lower end may be just fine.
JD Marlow
jd@jdmarlow.com
http://www.jdmarlow.com