Z-cam E2-M4 is worth a second look now that it does Pro Res internally at all frame rates including 4k @ 120fps, 4K 2.4:1 @ 160fps & 1080p @ 240fps.

[TomTheDP]

2 minutes ago, kye said:

It still needs an external display.  I have the Micro, but by the time you add an external monitor it's the size of a 1DX.

Also, I kind of don't really understand the logic of questioning why a camera needs to be small.  You can always add a rig if the camera isn't big enough.  I would love someone to direct me to the department for things I add to my rig to make the camera smaller / lighter.

The Sigma FP is a full-frame cinema camera that records RAW internally, and has decent battery life.  Really there's no excuses.

The issue with a lot of small cameras like the Sigma fp is the lack of a robust HDMI/SDI. Most serious stuff I do I'd need a monitor and I just can't trust the damn thing. 

I also find it easier to get a camera that is already 5-6 pounds for anything handheld. Rigging stuff is just a pain especially when you need to add 5-6 pounds to a half pound camera. IBIS is King though for run and gun stuff, it really just makes the job so much easier. 

Having a bit of heft also helps with IBIS. People say the S5 stabilization is not as good but in reality its because its lighter. The S1 gets more stable shots. When you get down to something the size of the sigma fp its a bit problematic for handheld which is what run and gun travel stuff often is. 

[Matt Kieley]

9 hours ago, Anaconda_ said:

What about the Z-cam E2c. It's the cheapest 4k raw option around, also internal ProRes, and about the size of the Blackmagic Micro.

I think All setting and monitoring can be done with a phone with no added hardware.

It'd be the closest thing to low cost, high quality crashcam... But is it low cost enough to purposefully smash it?

I used to have one. You only get prores and raw with an external SSD. And the Rolling shutter is pretty bad, even with fairly slow pans. And it's bad in both 4k and 1080p. Otherwise you can get a nice image out of it, with a lot of detail retained in the highlights.

[barefoot_dp]

I love my E2-S6 and am considering adding an E2-M4 as a b-cam for gimbal & long lens work.

[kye]

13 hours ago, TomTheDP said:

The issue with a lot of small cameras like the Sigma fp is the lack of a robust HDMI/SDI. Most serious stuff I do I'd need a monitor and I just can't trust the damn thing. 

I also find it easier to get a camera that is already 5-6 pounds for anything handheld. Rigging stuff is just a pain especially when you need to add 5-6 pounds to a half pound camera. IBIS is King though for run and gun stuff, it really just makes the job so much easier. 

Having a bit of heft also helps with IBIS. People say the S5 stabilization is not as good but in reality its because its lighter. The S1 gets more stable shots. When you get down to something the size of the sigma fp its a bit problematic for handheld which is what run and gun travel stuff often is. 

This is all true, but there are already 5-6 pound cameras!  I'm just saying I think there's a gap, not that every camera somewhere else in the features landscape should all be eliminated in order to fill my gap 🙂

The Micro has a full-sized HDMI.  In reality it's just a slightly larger connector, so having a robust socket isn't impossible.

[majoraxis]

On 3/13/2021 at 2:21 AM, IronFilm said:

Arguably didn't the GH5S / P4K / E2 / E2-M4 / Terra4K all share the same common origin as their underlying sensor they're derived/based upon?

I believe you are correct. So either the BRAW or variants of the IMX299 sensor may explain the the magenta shift when shooting under exposed on the Pocket 4k as compared to the E2-M4.

I am really enjoying the color science with the E2-M2 Flat color profile. I am able to get a good basic grade by using curves without a lut in Resolve.  I prefer not to use a lut if I don't have to.

[andrgl]

Can’t wait for their next gen camera. Literally only need to sort out 2 issues:

- External monitoring: lag, soft image, SDI-out in body.

- E-Mount that doesn’t suck. Viltrox nerfed the first iteration.

[majoraxis]

On 3/15/2021 at 11:42 AM, majoraxis said:

I am really enjoying the color science with the E2-M Flat color profile. I am able to get a good basic grade by using curves without a lut in Resolve.  I prefer not to use a lut if I don't have to.

I wanted to add that I am now using the Z CAM Color Correction OFX plugin for Resolve, which allows me to adjust the exposure and transform it to look like REC709.  I assume inside of the plugin, when you change the exposure, it applies the correct LUT based on the new exposure you set.  This fixes what I did not like about using LUTs, which is having them not matched correctly to the exposure and look bad in the under exposed or near over exposed areas.

The Z Cam Color Correction OFX plugin for Resolve as well as a version for FCPX, Z Cam LUTs and ZRAW plugin for Adobe Premiere and Scratch are available here: https://www.z-cam.com/lut/

[KnightsFan]

@majoraxisJust noting for those curious... The color plugin doesn't use LUT's, it uses color transforms. The nice part is it transforms to linear color before applying your adjustment, and then out to whatever you specify. So the white balance and exposure adjustments are actually correct, instead of incorrectly applied to non-linear data. Also by not using LUTs, it has none of the resolution or clipping problems that they would bring. The color plugin is another great reason to use Z Cam--outside of using Blackmagic color in Blackmagic's software, it's one of the best ways to maintain proper color management.

 

(Pro-tip: for any footage, convert to linear for WB adjustments! It's the reason why "raw allows WB adjustments"... really being Raw has nothing to do with it, it's just that it's linear so it's mathematically correct. That and no compression artifacts, in the case of lossless raw.)

[majoraxis]

@KnightsFan Thanks for the information - I really appreciate it!  I like the look of the ZCam plugin with the E2-M4 footage and I am thankful there is not a LUT involved.

The more I play with my Zcam ecosystem: ProRes 4:2:2, Flat (saturated log color profile), Z Cam Color Correction OFX plugin for Resolve with color transform to linear, Resolve Noise Reduction, Resolve OFX plugins such as Dehaze, the more I feel like it is completely useable without having to shoot ZRAW or switch cameras and shoot BRAW.

I think the E2-M4 is a worthy alternative to the Pocket 4k due to the results from the above work flow, as well as it beating the Pocket 4k in the other ways mentioned in my initial post. I would also like to add the with the 4k 120p slow motion comes lower rolling shutter, which is a benefit experienced in the low jello modes as well.

With some investment in learning how to get the most out of it, the value of the E2-M4 is very high compared to the competition.

[kye]

3 hours ago, KnightsFan said:

(Pro-tip: for any footage, convert to linear for WB adjustments! It's the reason why "raw allows WB adjustments"... really being Raw has nothing to do with it, it's just that it's linear so it's mathematically correct. That and no compression artifacts, in the case of lossless raw.)

I haven't heard this talked about much, so a couple of follow-ups if I may:

1. In Resolve, if someone does a CST to Linear, then adjusts WB, then CST back from Linear, which tool should be used for that WB adjustment?  I'm guessing it should be Offset?  The tools in Resolve all work in different gamma curves so I'm curious which one will work properly in Linear
2. In Resolve 17, they have changed the way the Temp and Tint sliders work so they now work correctly.  I saw this mentioned in passing in the middle of a colourists video, but the release notes don't say what they did, or why it's now correct, and no-one seems to be talking about it.  The colourist suggested that they used to be wrong and created artefacts, and it sounded like they might now work in Linear without you having to do it manually.  Any ideas?

[KnightsFan]

@kye

1. White balance is just linear gain on R G and B channels. I don't know the formulas for deriving Kelvin and Tint values. So after CST to linear, you could adjust gain values until it "looks right" and it should be correct though you don't know the Kelvin/Tint values--(if you're editing footage that doesn't have the original values in metadata, that wouldn't be possible anyway). Personally, I've found Resolve's WB values work as expected after a CST to linear. No idea if it's mathematically sound as proper linear gain at that point, but I did some basic tests comparing to in camera WB and it looked right to me. But now that I have an M4, I can just use WB sliders on the plugin and save some steps in the node graph as well.

2. I haven't heard that and am not sure how/if it changed. I've barely touched a physical camera since Covid broke out, so my Resolve 17 use has been solely for editing animations.

[kye]

4 hours ago, KnightsFan said:

@kye

1. White balance is just linear gain on R G and B channels. I don't know the formulas for deriving Kelvin and Tint values. So after CST to linear, you could adjust gain values until it "looks right" and it should be correct though you don't know the Kelvin/Tint values--(if you're editing footage that doesn't have the original values in metadata, that wouldn't be possible anyway). Personally, I've found Resolve's WB values work as expected after a CST to linear. No idea if it's mathematically sound as proper linear gain at that point, but I did some basic tests comparing to in camera WB and it looked right to me. But now that I have an M4, I can just use WB sliders on the plugin and save some steps in the node graph as well.

2. I haven't heard that and am not sure how/if it changed. I've barely touched a physical camera since Covid broke out, so my Resolve 17 use has been solely for editing animations.

The reason I bring up which controls is that I've heard that controls are designed to work in a colour space (for example, LGG in rec709) so they won't scale things correctly.  After a conversion to Linear, often the values are almost completely clipped, so I'm not sure what a control that's only meant to adjust things that are under (or close to) 100 IRE would do.  

I guess it's just a matter of knowing its an unknown and trying things if it doesn't look right.  I've definitely struggled with doing WB in post on non-RAW footage before, so it's useful to me to learn more and to hear that BM is working on improving it.

I just went back and found the reference to the WB mention in R17, and maybe I'm remembering it wrong, because maybe it's not more in-line with perception rather than just being purely mathematical, which is the next level again.  Perhaps they operate in XYZ colour space, or something similar?

Anyway, it's great that the ZCam OSX does the heavy lifting for you.  I keep hearing great things about Zcam from professional users, and it seems like they might be on track to grow and really mature as a brand and product lineup.

[KnightsFan]

9 hours ago, kye said:

After a conversion to Linear, often the values are almost completely clipped, so I'm not sure what a control that's only meant to adjust things that are under (or close to) 100 IRE would do.

If you do a CST, the values shouldn't clip. All of Resolve's processing is 32 bit float, so you only get clipping if you do an operation that has a hard ceiling or floor, like applying a LUT. I processed all my HLG XT3 footage with a CST to linear -> WB -> CST to Rec709 and never got any clipping. That is to say, it visually clipped in the monitor, but the data was still there are was operated on properly

Really you should never use a LUT for transforming color space unless you are forced to by poorly designed software. Even a 64x64x64 LUT has only 262,144 entries -- a frighteningly small number compared to 16 million 8-bit YUV colors. The lower resolution your adjustment has, the less you can do before showing artifacts. And going from Log or HLG to linear is a massive adjustment.

9 hours ago, kye said:

I just went back and found the reference to the WB mention in R17, and maybe I'm remembering it wrong, because maybe it's not more in-line with perception rather than just being purely mathematical, which is the next level again.  Perhaps they operate in XYZ colour space, or something similar?

Maybe? I guess if it looks good I don't really care if it's exactly precise. I just know that the typical method that some other software uses, which is to apply scalar gain to the gamma-encoded RGB values, looks really bad.

9 hours ago, kye said:

Anyway, it's great that the ZCam OSX does the heavy lifting for you.  I keep hearing great things about Zcam from professional users, and it seems like they might be on track to grow and really mature as a brand and product lineup.

I agree. I don't think they are "there yet" for general use by any working professional, but are certainly one of the most exciting camera companies to watch right now. They have little problems, like needing 2 dongles for timecode, no SDI, weird audio clipping problems (possibly fixed in firmware? I haven't tested), and messy menus. But their core product is so far ahead of any of their competition, many people are willing to put up with it. I'm sure their next generation will be great.

[majoraxis]

The video touches on the topic of exceeding the color space and using the a scope and a plug in Resolve to monitor it and keep it within REC709 limits. The section starts at 14:52.

Use the CIE Chromaticity Scope in Resolve to view the color space and how your levels exist inside and outside of it.

If and when you are exceeding it, add the Color Space Transform Resolve OFX plugin and select the Gamut Mapping option in the plugin. Set it to Saturation Mapping and adjusting the Saturation Max and Knee to your color space target max levels.

The video also has some good tips on grading Zcam footage, but can be applied in general.