Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Forums DaVinci Resolve Vectorscope accuracy once and for all! COMPARISON

  • Vectorscope accuracy once and for all! COMPARISON

     Chris Oben updated 9 years, 7 months ago 3 Members · 4 Posts
  • Hampus Lager

    February 9, 2012 at 10:27 pm

    Hi all!

    A few days back I had to calibrate a monitor (not a grading monitor) without an i1 or the like for a friend. He mostly do web and sometimes DVDs (not high paying jobs, he’s just enjoying his hobby). So I brought some test patterns and fired up Resolve Lite and outputted via HDMI (Decklink Studio 2) to the monitor. I loaded the test patterns and I was shocked to see that the internal vectorscope was off. I tried other bars as well (got loads of them), going between ACES and DaVinci YRGB with no difference. I fired up Resolve on my MBP and loaded the bars and it was the same thing there. I checked them in Scopebox 2.0 as well. Off as well. Something wrong with the patterns? Fired up Avid and looked at that vectorscope: Perfect. After Effects and Color Finesse: Pretty much perfect.

    This has driving me nuts the last few days. I’ve found nothing online, only more unanswered questions. I grabbed a Sony EX1r and tested the in-built color bars via both HDMI and SDI in Scopebox (Decklink Studio 2). Vectorscope was off there as well. Recorded the signal i Media Express and played back in Scopebox. No difference except the compression (Prores HQ).

    This got me questioning everything. It can’t be a bug in both Scopebox and Resolve. It can’t be that all patterns and bars are wrong. It can’t be all the different computers I’ve tested this on that are wrong. I checked them all with the color picker in After Effects and made sure that the colors are 75% and correct (192 in 256 steps in 8-bit, just to make the math simple) and yes they are. I want to be a colourist and I got no problem with the tools used to grade or aesthetic choices. But I’m also a tech geek and I need to understand this in order to sleep at night.

    The patterns I’ve used (Prores) can be found in my dropbox: https://dl.dropbox.com/u/17224404/scopes.zip

    Anyone got anything on this?

    SMPTE https://i40.tinypic.com/655×68.jpg

    Belle Nuit: https://i43.tinypic.com/syseaa.jpg

    DVE HD Basics: https://i41.tinypic.com/fvbhja.jpg

    Sony EX1r: https://i44.tinypic.com/f3dfzs.jpg

  • Alexis Hurkman

    February 12, 2012 at 4:45 pm

    I’ve loaded all three of your test pattern files onto my Resolve suite machine, and output to my Harris VTM 4100 video scope. The Belle Nuit pattern is off, it’s displaying 100% and 75% bars, but it looks like the file was generated scaled to full digital range, rather then studio range, so that explains one of your issues.

    However, the other two test pattern files (SMPTE and DVE) fall right into the Harris’ target boxes. The DVE pattern’s Cyan bar is hugging the inner-left corner a bit, but it’s still inside.

    Comparing this output to the DaVinci software scopes, (a) it’s hard to read because the DaVinci scopes show a scatter not a trace, (b) they appear to be off for some targets but not others. If I were a betting man, I’d say the graphs themselves are probably accurate in shape, but the DaVinci software vectorscope targets are possibly being drawn for the wrong color space for some reason. However, that’s just a guess.

    Using the latest beta of Scopebox, the SMPTE and DVE patterns fall into the proper boxes of the Vectorscopes, so you may be seeing an issue from version 2 that will be fixed when version 3 comes along. Either that, or there’s a color space setting that needs to be manually adjusted in version 2 (I don’t know Scopebox all that well).

    http://www.alexisvanhurkman.com | http://www.correctionforcolor.com

  • Hampus Lager

    February 12, 2012 at 5:00 pm

    Thanks Alexis 🙂

  • Chris Oben

    January 8, 2013 at 6:13 am

    Hi Hampus,

    I’m dealing with similar issues and wondered if you could post your test files to dropbox again…

    C.

    Chris M. Oben

    https://www.chrisoben.com

Viewing 1 - 4 of 4 posts

Log in to reply.

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy