GH4 Shane Hurlbut tests V-log on the GH4

Thomas Smet

Veteran
Shane just released an interesting test of V-log on the GH4.

https://www.hurlbutvisuals.com/blog...log-l-latitude-test-cinematography-education/

Its interesting that he suggests under exposing by a single stop. That sort of goes against a lot of the information of ETTR. It may make sense for color but I think at the expense of shadow detail. It is nice to hear him have some nice words for the Gh4 and it sounds like he feels V-log is a huge update to the Gh4. Wonder how he would feel about Leeming LUT?
 
Interesting. The skin tones look more even and the colour is more appealing. Yellow versus red.

There seems to be more red in her top though as well so has the colour correction been done so the charts match?
 
this is confusing. he's using his own set of luts, not the Panny one that was designed for the GH4 VLog, so THOSE luts work better with vlog at -1?
 
I'm going to go out on a limb here and say that his tests don't look natural or nice at all, and that he is touting settings such as underexposure that are going to give people more noise from the camera. wwjd is right - this will end up adding a lot of unnecessary confusion to the community.

The model's skin looks quite yellow for example. He mentions that he is using a creative LUT that fits his "Into The Badlands" style, so the testing is already swaying away from what we should be looking for as a baseline.

I'm in the process of preparing a couple of short video tutorials and tests to showcase my LUT and how it gives you the best possible image from the GH4. I think I might see if I can get a nice model to help out, as a kind of counterpoint to Shane's video, then see which one people prefer.

I should be able to do it at 10bit too, by borrowing an external recorder, to give V-Log a good run as well.

Stay tuned! :D

Cheers,

Paul :)
 
To be fair I think Shane was mostly focused on skin tones and nothing else and he tests cameras to fit within his workflow. His look has been successful for him so I'm not sure how it can cause confusion, its just another method. Shane is pretty detailed in his tests and has explored this subject in greater detail than most tests of the GH4 have.

To each their own when it comes to a look.
 
To be fair I think Shane was mostly focused on skin tones and nothing else and he tests cameras to fit within his workflow. His look has been successful for him so I'm not sure how it can cause confusion, its just another method. Shane is pretty detailed in his tests and has explored this subject in greater detail than most tests of the GH4 have.

To each their own when it comes to a look.
I mean confusion more in the sense of people latching onto the "expose V-Log at -1 for best exposure" kind of simplistic soundbite, without understanding why. I guarantee you'll see someone say that somewhere and point to Shane's name and status to say that it must be right.

Looks are definitely a personal thing, I agree :)
 
I mean confusion more in the sense of people latching onto the "expose V-Log at -1 for best exposure" kind of simplistic soundbite, without understanding why. I guarantee you'll see someone say that somewhere and point to Shane's name and status to say that it must be right.

Looks are definitely a personal thing, I agree :)

Completely agree. I'm not convinced -1 is the way to go either and it may work well for his look but may not work the same for other looks. Shane is planning on selling his LUT for the GH4 so for those that do like his look it is a thing to consider. I of course like to explore a lot of options. I may try his LUT and -1 method but I will keep an open mind and explore other LUTS as well. I haven't dived into V-log yet so its a bit of a non issue for me right now anyway. For now I am using Leeming and ETTR because that combo works magic. I will not use V-log until I can buy an external 4k recorder. I tried external HD recording but I'm not convinced the down scaling in camera from 4k to HD is of the best quality and want to wait until I can afford an external 4k recorder.

Besides I think it would be good to see how well this works under different lighting situations. Not all of us can afford $25,000 of lights and a $800 light meter and a crew of cult members.
 
I don't yet have V-Log for my GH4, but at my current job I have been using Sony cameras, S-Log, etc. so I am educating myself about Log shooting, LUTs, etc.
What strikes me as strange about Mr. Hurlbut's recent tests are the extreme color shifts whenever he lifts or lowers exposure to compensate for his intentional over- or underexposure. A simple raising or lowering of exposure shouldn't radically alter the tint of the image, and yet his does, which makes me very skeptical of his LUT. Or perhaps he is not being transparent about just what he and his colorist are doing. He also mentions that his colorist couldn't add saturation without it oversaturating. Which again makes me skeptical of that LUT. The model's hair going unnaturally yellow in overexposure is a clue, as the original V-Log image looks fine.

All the more reason to perform your own tests and see what works for you.

I agree though that this has the potential to just cause more confusion, as Shane Hurlbut is well-known in this community, and some may indiscriminately underexpose their images with unsatisfying results.

*EDIT I took a cue from Thomas Smet and applied the official Panasonic V-Log LUT to screen grabs of Hurlbut's ungraded V-LOG images. The image tints and shifts are very similar. Interesting.
 
Last edited:
I agree, whatever he wants to do is cool and all, but without KNOWING what his luts are doing, the -1 is a tad pointless to us users. It does work for him, but not for us. Good testing though.
 
I'm currently unable to see footage etc. Just throwing this out there: underexposing V-LOG L makes a lot of sense. Overexposing it, however, borders on insane. :)

V-LOG L is an asymmetrical curve. It holds 8 stops below middle gray, and 4 stops above. If you underexpose by a stop, you effectively get 7 stops below and 5 stops above. That is closer to the original intent of the original VLOG (which was designed as 8 below / 8 above).

Underexposing will increase noise, yes (when you gain up the mids and shadows to proper exposure). But it will retain the 7/5 split.

On the other hand, overexposing results in a grotesque disparity in stop apportionment. Follow some S-Log shooter's advice and overexpose 2 stops and you'll end up with 10 stops below middle gray and only 2 stops above! That is absurd. In a case like that you'd have only maybe one stop from proper skin exposure to pure white blowout. And on a GH4 it is even potentially worse because its sensor doesn't deliver a full 12 stops.

I would much rather have some headroom over the skin tones, which is what a little underexposure can provide. Yes I know that S-log and C-log shooters frequently call for a whole lot of overexposure, but they're not using V-Log L. I say run your own tests, especially daylight exteriors with bright objects and skin, and see whether you agree that the idea of some underexposure has merit for V-Log L.
 
Last edited:
V-LOG L is an asymmetrical curve. It holds 8 stops below middle gray, and 4 stops above. If you underexpose by a stop, you effectively get 7 stops below and 5 stops above. That is closer to the original intent of the original VLOG (which was designed as 8 below / 8 above).

I think you mean "6 below / 6 above", right?
 
I think you mean "6 below / 6 above", right?

The GH4 V-LOG L is a subset of VARICAM V-LOG which is designed for 8 stops below middle grey and 8 stops above middle grey. The GH4 version of VARICAM V-LOG is 8 stops below and 4 stops above, so under exposing by one stop you sacrifice 1 stop in the shadows but gain 1 stop in the highlights.
 
I'm currently unable to see footage etc. Just throwing this out there: underexposing V-LOG L makes a lot of sense..

Even without V-LOG, and just using the Natural profile, I have had good results underexposing just a tad, maybe half or 2/3 stop, when shooting at 400 ISO and raising it up in post. Skin looks great. Highlights look decent.
 
I agree, whatever he wants to do is cool and all, but without KNOWING what his luts are doing, the -1 is a tad pointless to us users. It does work for him, but not for us. Good testing though.

to be fair I took some screen grabs of his video of the LOG shots and used the Panasonic LUT on them and I am seeing pretty much the same results. Yes the 0 and +1 have less noise and overall look cleaner but that is at the expense of color. The -1 even with the Panasonic LUT has richer colors. It comes down to a tradeoff of noise vs color and pick your poison. The 0 exposure does look very nice but the color is definitely lacking compared to the -1 exposure.
 
is this "richness", simply saturation based.... colored to taste? by his LUT?

I guess I'm less concerned with "out of camera" since everything gets corrected/graded anyway. I don't expect a $1400 camera to equal Alexa out of camera, but with some grading, most viewers could never tell the difference
 
is this "richness", simply saturation based.... colored to taste? by his LUT?

I guess I'm less concerned with "out of camera" since everything gets corrected/graded anyway. I don't expect a $1400 camera to equal Alexa out of camera, but with some grading, most viewers could never tell the difference

This has nothing to do with his LUT. I used the Panasonic LUT with his material. Well a screen grab of his video of the LOG shots. I wanted to see if maybe his LUT was doing something funky and the Panasonic LUT showed the same thing. This is really about color. I think we tend to focus on ETTR because we strive to get the cleanest image possible out of the GH4 with as little noise as possible but that process may in fact sacrifice color. The point of all of this really is just how much room can you give yourself to adjust color and highlights in post. Like Barry said, ETTR gives lots of room to play with the shadows but the by product is virtually no room to play with the highlights.
 
This has nothing to do with his LUT. I used the Panasonic LUT with his material.

You are absolutely right. I just did the same thing. So weird to see how, at 2 stops overexposed, the yellow just goes insane. Not that I would have wanted to overexpose V-Log by 2 stops, but it's interesting to see.
1 stop underexposed does look good, but it is definitely noisier than the image at nominal exposure. The V-Log LUT does shift the tint of the image but not as extremely as Hurlbut's LUT.
I'm still concerned about the way V-Log creates magenta blocks in the grays and shadows.
 
Back
Top