mercer Posted November 29, 2018 Share Posted November 29, 2018 Good info, thanks @frontfocus!!! I didn’t know FLog was 2020. Is anyone experiencing any color shifts with a Rec709 transform? Quote Link to comment Share on other sites More sharing options...
DanielVranic Posted November 29, 2018 Share Posted November 29, 2018 35 minutes ago, mercer said: Good info, thanks @frontfocus!!! I didn’t know FLog was 2020. Is anyone experiencing any color shifts with a Rec709 transform? Nope - I actively avoid the Eterna LUT - really burns the footage. I use the Fuji Flog-709 LUT and I love it. mercer 1 Quote Link to comment Share on other sites More sharing options...
BrunoCH Posted November 29, 2018 Share Posted November 29, 2018 3 hours ago, mercer said: I’m curious why there are noise issues with the 2K if there isn’t in 4K? Is it common practice to ETTR F-Log? Does anyone know where middle grey is at in F-Log? Me too I am curious to know. If I find the time, tomorrow or Saturday I do a test 2K vs 4K about moiré and noise. I'll share the file on google drive. About exposing F-log, when I shoot a scene with a low contrast (very compress signal), I expose to the right. When I shoot a scene with high contrast, I expose to preserve the highlights. mercer 1 Quote Link to comment Share on other sites More sharing options...
mercer Posted November 29, 2018 Share Posted November 29, 2018 21 minutes ago, DanielVranic said: Nope - I actively avoid the Eterna LUT - really burns the footage. I use the Fuji Flog-709 LUT and I love it. So the Eterna LUT is a utility LUT and affects exposure? Or do you mean it is burning saturation and a look into the image. Are you exposing to the right before using the FLog-709 LUT? 21 minutes ago, BrunoCH said: Me too I am curious to know. If I find the time, tomorrow or Saturday I do a test 2K vs 4K about moiré and noise. I'll share the file on google drive. About exposing F-log, when I shoot a scene with a low contrast (very compress signal), I expose to the right. When I shoot a scene with high contrast, I expose to preserve the highlights. Great info thanks, I look forward to your tests. Quote Link to comment Share on other sites More sharing options...
BrunoCH Posted November 29, 2018 Share Posted November 29, 2018 1 hour ago, mercer said: Good info, thanks @frontfocus!!! I didn’t know FLog was 2020. Is anyone experiencing any color shifts with a Rec709 transform? On the file I shared above, the green of the pullover is not really accurate. But with 10 bits, there is so much latitude to correct. Quote Link to comment Share on other sites More sharing options...
DanielVranic Posted November 29, 2018 Share Posted November 29, 2018 1 hour ago, mercer said: So the Eterna LUT is a utility LUT and affects exposure? Or do you mean it is burning saturation and a look into the image. Are you exposing to the right before using the FLog-709 LUT? Great info thanks, I look forward to your tests. I will post some screenshots tonight to show you. Also. With 10Bit, I stopped exposing to the right for log. I like my highlights. webrunner5 1 Quote Link to comment Share on other sites More sharing options...
DanielVranic Posted November 29, 2018 Share Posted November 29, 2018 Also I did 2K vs 4K tests for a while last night -- I will also post those results Quote Link to comment Share on other sites More sharing options...
thebrothersthre3 Posted November 30, 2018 Share Posted November 30, 2018 If I need high ISO's I try to use 4k otherwise I think 2k is great. I never reframe in post though. Quote Link to comment Share on other sites More sharing options...
DanielVranic Posted November 30, 2018 Share Posted November 30, 2018 Attached are my examples - TIFF exports from Resolve- First 2 are 2020-709 conversion, 2K and then 4K - 2nd set is Eterna LUT 2K and 4K Eterna4K.tif Eterna2K.tif 2020-7092K00087016.tif 2020-7094K.tif Quote Link to comment Share on other sites More sharing options...
IronFilm Posted November 30, 2018 Share Posted November 30, 2018 On 11/28/2018 at 6:05 AM, Kisaha said: And 2.8f, and no IS. Not the modern workhorse lens that everyone else has. For many people the 24-105mm f4 is their "workhorse lens", which is F4 On 11/28/2018 at 9:58 PM, austinchimp said: I think there's also a bias in the kind of footage we've seen from both cameras. The Pocket seems to have been bought by more indie filmmaker types, while the Fuji seems to be in the hands of more photographers from what I can tell by going on YT and Vimeo. So I think the Pocket footage is appealing if you match that profile. A lot of the Fuji footage is badly shot, exposed and graded. That first party is key. The type of people shooting with a camera will very heavily influence what we think a camera "looks like" Quote Link to comment Share on other sites More sharing options...
Kisaha Posted November 30, 2018 Share Posted November 30, 2018 6 minutes ago, IronFilm said: For many people the 24-105mm f4 is their "workhorse lens", which is F4 In APS-C is 2.8f for sure. The Samsung NX is 2-2.8f. For full frame systems 4f is probably good enough for most things. Quote Link to comment Share on other sites More sharing options...
thephoenix Posted November 30, 2018 Share Posted November 30, 2018 f4 is the same no matter the sensor size Quote Link to comment Share on other sites More sharing options...
BrunoCH Posted November 30, 2018 Share Posted November 30, 2018 10 hours ago, DanielVranic said: Attached are my examples - TIFF exports from Resolve- First 2 are 2020-709 conversion, 2K and then 4K - 2nd set is Eterna LUT 2K and 4K Eterna4K.tif Eterna2K.tif 2020-7092K00087016.tif 2020-7094K.tif Your 2Ktiff is a 3840x2160 picture !?! Quote Link to comment Share on other sites More sharing options...
DanielVranic Posted November 30, 2018 Share Posted November 30, 2018 1 hour ago, BrunoCH said: Your 2Ktiff is a 3840x2160 picture !?! Crap - I will try again~ Quote Link to comment Share on other sites More sharing options...
BrunoCH Posted December 1, 2018 Share Posted December 1, 2018 I did a quick 2K vs 4K test. You can download "trimmed" copies into QT of the original files. Only the length changes from the originals. No transcoding. For the first two files (79 and 80) I wanted to compare moiré. But I didn't reproduce moiré (it is however the same pullover I've get moiré). Just a little bit of moiré above the bottom button on the 2K file. Nothing on the 4K. Not enough to conclude that there is less moire in 4K. I will try to repeat a test in another way. For files 82 and 83 I wanted to compare the noise. There is a little less noise on the 4K vs 2K in a 2K timeline in FCPX. All files are F-Log H265 10bits All-I, 400Mbps for 4k and 200Mbps for 2K. 79 and 80 : 680 ISO exposed to preserve the highlights. 82 and 83 : 3200 ISO exposed on the right. (82 and 83 are out of focus. It is very difficult to focus on a Log ETTR) https://drive.google.com/open?id=17e1c60bj81n8BQwoLP9xtQKKhCgOkC6a Quote Link to comment Share on other sites More sharing options...
DanielVranic Posted December 3, 2018 Share Posted December 3, 2018 Did many more 2K vs 4K test this weekend. 4K wins by a huge margin every time. Noise is much uglier and footage gets really unusable past 5000 ISO. 4K is still OK. Still trying to find a workflow to convert to 709. Sony was much easier in Resolve to do Color Space Transform. Fuji doesn’t have an exact curve in their system. Had best luck so far using Color Space Transform and setting it to Cineon Log and Rec.2020. My second technique was using the Fuji 2020-709 LUT and then doing a manual S-Curve on multiple nodes If anyone know a better way, please let me know. Quote Link to comment Share on other sites More sharing options...
androidlad Posted December 3, 2018 Share Posted December 3, 2018 2K/HD on X-T3 is derived from weighted horizontal pixel binning and vertical line skipping, it's one of the main features of the 3.76um generation Sony sensor, it's an on-chip operation so less work for Fuji and less processing requirement. Quote Link to comment Share on other sites More sharing options...
BrunoCH Posted December 3, 2018 Share Posted December 3, 2018 17 hours ago, DanielVranic said: Did many more 2K vs 4K test this weekend. 4K wins by a huge margin every time. Noise is much uglier and footage gets really unusable past 5000 ISO. 4K is still OK. Still trying to find a workflow to convert to 709. Sony was much easier in Resolve to do Color Space Transform. Fuji doesn’t have an exact curve in their system. Had best luck so far using Color Space Transform and setting it to Cineon Log and Rec.2020. My second technique was using the Fuji 2020-709 LUT and then doing a manual S-Curve on multiple nodes If anyone know a better way, please let me know. For my part, the noise level difference 4K vs 2K isn’t so important to making me change workflow. Remember human perception is more sensitive to noise level change than noise itself when it is at a reasonable level. I don’t use ISO sensitivity as an exposure parameter (because the sensor only has one sensitivity actually). When I shoot, most of the time, I use two sensitivity (one for exteriors and one for interiors) to maintain a constant noise level. I use FCPX for editing and color grading. 11 hours ago, androidlad said: 2K/HD on X-T3 is derived from weighted horizontal pixel binning and vertical line skipping, it's one of the main features of the 3.76um generation Sony sensor, it's an on-chip operation so less work for Fuji and less processing requirement. Thank you for these explanations. Do you have links to share? And for 4k so how's it going? Quote Link to comment Share on other sites More sharing options...
androidlad Posted December 3, 2018 Share Posted December 3, 2018 7 minutes ago, BrunoCH said: For my part, the noise level difference 4K vs 2K isn’t so important to making me change workflow. Remember human perception is more sensitive to noise level change than noise itself when it is at a reasonable level. I don’t use ISO sensitivity as an exposure parameter (because the sensor only has one sensitivity actually). When I shoot, most of the time, I use two sensitivity (one for exteriors and one for interiors) to maintain a constant noise level. I use FCPX for editing and color grading. Thank you for these explanations. Do you have links to share? And for 4k so how's it going? 4K is super-sampled from 6240 x 3510 (<30p), 5266 x 2962 (60p). Quote Link to comment Share on other sites More sharing options...
BrunoCH Posted December 3, 2018 Share Posted December 3, 2018 16 minutes ago, androidlad said: 4K is super-sampled from 6240 x 3510 (<30p), 5266 x 2962 (60p). What do this mean? 6240x3510 video out of sensor downscale in 4K? Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.