GH4 Adam Wilt of PVC exposes the facts about V-Log L

What's most interesting to me is how the DVX200 LOG outperforms the GH4 LOG, as I had always assumed that they were pretty much the same.

Hopefully the GH5 will perform a lot better when it comes to shooting with a LOG profile.
 
TheDingo said:
What's most interesting to me is how the DVX200 LOG outperforms the GH4 LOG, as I had always assumed that they were pretty much the same.
Click to expand...
Lots of people continue to assume that, but they're just not -- the DVX200 has the next-generation sensor in it, and its processing engines are more powerful. I don't have a GH4 so I couldn't compare directly, but when we had the great VLOG wars of 2015, when people were absolutely losing their minds over it, I kept trying things on the DVX200 and saying "where's the problem?" I don't think the problem is in VLOG, it's in the inherent limitations of a $900 camera body, as Adam pointed out. It will be interesting to see how the situation fares in the next generation of camera (presumably a GH5 someday).
 
DLD said:
That's what I thought but they surely seem to be a lot of bars on that chart.

(maybe I've run out of fingers and toes to use)
Click to expand...

Well remember to look at the Waveform, you'll see you have to start counting at the third bar to count properly as the first one is clipped and the second one just clip or just under clip and represents stop 0. I count at most 12-13 (with 13 being barely visible even with a lot of manipulation) for DVX200 and 11-12 for GH4. By comparison similar xyla chart tests seem to show about 15-16 for A7SII and 17-18 for Alexa/Dragon/Varicam etc.
 
"...while the GH4 uses the 16-235 range for its internal recording."

Is that really right, though? The GH4 locks the setting at 0-255 when in V-Log. The user can't choose the 16-235 normally available for other picture profiles.

One thing I notice is that I'm using firmware 2.5, and Adam was using 2.3. However, I saw no mention of a change on this front with the 2.5 update.
 
Barry_Green said:
Lots of people continue to assume that, but they're just not -- the DVX200 has the next-generation sensor in it, and its processing engines are more powerful. I don't have a GH4 so I couldn't compare directly, but when we had the great VLOG wars of 2015, when people were absolutely losing their minds over it, I kept trying things on the DVX200 and saying "where's the problem?" I don't think the problem is in VLOG, it's in the inherent limitations of a $900 camera body, as Adam pointed out. It will be interesting to see how the situation fares in the next generation of camera (presumably a GH5 someday).
Click to expand...

I'm still convinced the issues with V-log on the GH4 have everything to do with the h264 encoder and nothing else. While the encoder is great in many situations it groups similar colors together way too aggressively. That is a major problem with a log format where a lot of colors are similar. In most cases even recording external at 8bit works great which goes to show it is really the h264 encoder that just doesn't handle V-log the best.
 
I'm also not entirely sure I agree with using luts for V-log. Sure it's an easy solution to get a decent grade but to me that entirely defeats the whole point to shooting log which is a sort of pseudo raw format.

Log is all about having less of a baked in look and having more range to dial in a look in post. It is about being able to make a decision in post if one wants to sacrifice the lows or highs instead of making that decision while shooting. We shoot flat because we want to fineness the look in post and not have a canned look out of camera. Slapping on the Panasonic LUT isn't much different than just shooting a regular profile on the GH4.

I love V-log for the massive headroom it gives me to grade manually. V-log 10bit really does act like a poor mans raw format.
 
Thomas Smet said:
the whole point to shooting log which is a sort of pseudo raw format.
Click to expand...

I don't think that's even a little true. Log and RAW are not analogues in any way.

They are both used to achieve similar end results - maximum dynamic range. But the methods they use to do that are so vastly different that any comparison only invites misunderstanding, I think.
 
Jim Simon said:
I don't think that's even a little true. Log and RAW are not analogues in any way.

They are both used to achieve similar end results - maximum dynamic range. But the methods they use to do that are so vastly different that any comparison only invites misunderstanding, I think.
Click to expand...

I disagree. Even Charles Ponyton the famous color scientist has stated that 10bit log gets you 99% of what is actually usable from 12 bit raw. Some information is lost such as 4:4:4 color and the ability to de bayer the way you want but in terms of grading you are not missing much at all.

Of course log is not the same as raw. Two totally different things but in terms of grading they both give the same benefits with the log typically easier and less system intensive to work with. Log is more of a baked in profile of course and requires more precise exposure and white balance in camera than what one can get away with raw but I don't consider that a big hinderence of log since I want to exposure and white balance my shots properly anyway. This is why I said a pseudo raw format and not the same as raw.
 
Thomas Smet said:
in terms of grading they both give the same benefits
Click to expand...

I will agree that Log is easier for most of us to work with, and that like RAW files we have more freedom in post with Log than we do with other profiles (though not nearly as much as we get with RAW). But still...Log isn't RAW, not even a pseudo version. And I still think that your comparison only invites misunderstanding.

Log is a baked in profile. RAW is a recording format (which may or may not be Log itself). Two very different things.
 
You must log in or register to reply here.
Back
Top