Sigma Fp review and interview / Cinema DNG RAW

[mercer]

2 hours ago, Lars Steenhoff said:

I downloaded the shot with the girl next to the water and when I push the exposure a bit you can clearly see moire in the dress around her hip.

good thing I plan on using some vintage lenses that are not that clinically sharp.

 

I'm happy with highlight recovery in resolve, it gets back some very hot spot on her arm.

 

 

If you look at the shadowed part of her dress, it looks like chroma noise to me. You'll see it with ML Raw with underexposed shadows, but this shot is hardly underexposed. Hmm.

[Lars Steenhoff]

It's moire, Its more obvious when playing the whole clip. 

The dress has fine lines that create this effect.

https://www.dropbox.com/s/newx1n59ihz0lbl/moire.mov?dl=0

Here' the full clip, I developed it with sharping quite high because I wanted to see what was going on

I did notice that when I open the files in adobe camera raw the moire effect is almost gone, so it may also be something todo the way resolve handles this

[paulinventome]

The charts and the sharpness tests are tough metrics and not real world either.

AFAIK the downsampling is in some way to minimise moire because there is no OLPF. You can't really have an OLPF in a camera that switches modes or does stills. I've seen plenty of shots ruined because of that from some quite well known 'prosumer' cameras as well. And BMD is quite notorious for this. They score well on tests but not so much in real world.

But the sigma image is quite soft. 

What the slashcam shows though is a difference between crop and normal and that's interesting. So full frame is downsampled but then it's also fair to say the downsampling isn't an even number and maybe that's why we're seeing these kind of results whereas the S35 mode might actually be putting the sensor into a windowed mode and that's worth looking at more.

The 12 bit footage - i strongly think if you are recording 12 bit linear to push the image as high as you can and bring down. Because that really is linear, the brightest stop of light will show nearly 2000 steps of gradation. So push as far as you can to take advantage of the tonality. Of course that's not always possible. The 8 bit footage less so because they are reshaping the data so that top stop takes much less space. Again i'd really hope for 10 shaped bit from 12 bit linear.

I do believe if this is the IMX sensor then there is a 4K 14bit crop mode which would be awesome.

But with no OLPF then as an overall format you'd see much more aliasing.

Lightroom is quite a bad debayer - i see artefacts all over the palace. Resolve is a bit better. SpeedGrade was poor. And nuke used to have some really good options in there (but those got 'upgraded')

Motion RAW is quite a bit more difficult than stills RAW and i don't think there is a good DNG option for debayering out there. It would be possible to use temporal debayer but i am no aware of anything like that which exists...

Also the reason why it seems better to push the shadows that bring the highlights down is something that just occurred to me on a long drive back. The 12 bit is just truncating the top stops. The whole camera exposure is the same for stills vs cine. So same settings like for like, but the cine has the 2 bits chopped off. So therefore the expectation is for there to be highlights compared to mid grey and it's not quite there. That 14 bit 4K mode would be perfect!

cheers
Paul

[rawshooter]

Speaking of exposure, I can't get the zebras to work on my fp (with firmware 1.01). They appear in the menu, but are no longer displayed after leaving the menu.

Is there some trick to get this to work?

[rawshooter]

Figured it out with the help of the camera manual. The menu entries for the zebras are only for configuring their parameters, not for activating them. For the latter, one needs to define a custom shoot mode and then customize the display settings for that shoot mode (including activation of zebras).

Zebras and focus peaking don't work together. And in Cine mode + Manual mode, lens aperture cannot be controlled via the camera's wheels, but only via menus, which is a huge downer for lenses without aperture rings (such as adapted ef mount lenses).

Many configuration items - such as audio level adjustment - are buried in submenus.

In still mode and M setting (with Auto-ISO), exposure compensation is only available via menus, not via dials - another huge letdown. The camera tends to overexpose in automatic metering and has no highlight priority metering option. 

This is definitely not a straightforward camera in handling and everyday use.

 

[paulinventome]

On 12/1/2019 at 4:15 PM, mercer said:

If you look at the shadowed part of her dress, it looks like chroma noise to me. You'll see it with ML Raw with underexposed shadows, but this shot is hardly underexposed. Hmm.

IMHO most likely crappy debayering combined with some serious 'clarity' adjustments in that last shot.

Whilst i'm all for destruction testing images i also have to balance with i don't shoot like that. I'm more interested in the things that do show all the time, tonality and range. 

If you ever get to push lightroom around on high contrast edges check some really odd debayering artefacts - and this is supposed to be pro-stills level.

The beauty of debayering is the ability to choose different algorithms depending on scene - of course no one actually offers that...

cheers
Paul

[Lars Steenhoff]

You can try this app

https://mlv.app

You still need to convert first the dng to mlv

https://www.magiclantern.fm/forum/index.php?topic=24631.0

It allows you to choose different debayering methods.

 

[Lars Steenhoff]

Any SD card recommendations?

I need to record 8 bit 4k 25 fps internal.

The largest I can find from the recommended list is the 128 gb sandisk uhs2 

I would prefer a larger card, anyone tested other brands? like pro grade

[Brian Williams]

9 hours ago, Lars Steenhoff said:

Any SD card recommendations?

I need to record 8 bit 4k 25 fps internal.

The largest I can find from the recommended list is the 128 gb sandisk uhs2 

I would prefer a larger card, anyone tested other brands? like pro grade

I tried the Lexar 1667x, but as expected it was too slow (was thinking that 250MB/s was write speed, not read). Thinking the 2000x would work, but haven’t seen it bigger than 128gb.

They’re all sooo expensive.

[paulinventome]

12 hours ago, Lars Steenhoff said:

I would prefer a larger card, anyone tested other brands? like pro grade

I got some progrades off amazon the 128gb version and has worked no problem. Was a bit more cost effective than the others. I think they benchmarked higher than others and are rated 300MB/s (which is above 25p at 8 bit)

cheers
Paul

[rawshooter]

Just saw this:

https://docs.google.com/spreadsheets/d/1FUt1bfhWYftUhlABkO91goSNEYuPZzlrxmOVGDJnQ7E/edit#gid=0

A collaborative Google doc for tracking current issues of the Sigma fp and submitting them as a bug report to Sigma.

 

[Lars Steenhoff]

Here's my list:

 

• allow af-on butwton when in manual focus mode
• Allow 2 second timer for movies. To disable shake at the beginning of the shot when on tripod
• Allow manual focus override when in autofocus mode
• let us change focus direction 
• linear mode for focussing
• magnify during movie recording
• magnify time after touch lens 0.5 second
• LUT support / Bake LUT in camera
• zebra and focus peaking at the same time
• 10 bit non linear raw internal
• non linear raw 10 bit external
• lossless compressed raw
• 12 bit raw at 25 frames per second

[MikhailA]

5 hours ago, rawshooter said:

Just saw this:

https://docs.google.com/spreadsheets/d/1FUt1bfhWYftUhlABkO91goSNEYuPZzlrxmOVGDJnQ7E/edit#gid=0

A collaborative Google doc for tracking current issues of the Sigma fp and submitting them as a bug report to Sigma.

 

 

"after powering off the camera, or after power save sleep, the camera loses basic settings such as video resolution, codec, audio recording level"  - is that true?

[rawshooter]

32 minutes ago, MikhailA said:

 

"after powering off the camera, or after power save sleep, the camera loses basic settings such as video resolution, codec, audio recording level"  - is that true?

Unfortunately yes.

[paulinventome]

2 hours ago, rawshooter said:

Unfortunately yes.

Mine doesn't... Or at least i've not noticed this. 

Can you give me a specific setting that you lose after power off?

Mine keeps recording format, ISO, shutter, etc. What about timecode? Mine keeps timecode running too... I'm not sure how accurate the TC is, i ought to test with some tentacles but on the face of it i am not loosing anything. I'd go crazy with frustration if it did!

cheers
Paul

[paulinventome]

As an aside, i did a quick bit of comparison with an SSD plugged in now. I think it was pretty expected:

12 bit DNGs, the best quality both in Shadows and Highlights - the shadows are also not biased green
10 bit DNGs, these are identical to 8 bit in the shadows, band for band. However the highlights roll off with more data in so they match the 12 bit and work with highlight reconstruction better
8 bit same shadow performance as the 10 bit and the highlights reach the same overall range but the values in the highlights are less tonal and this yields magenta tints. So highlight reconstruction not so quite as good as the others.

In all cases i found it near impossible to 'break' the midtones

I wonder whether that linearisation curve in the 8 bit is a little off with respect to each of the channels, with green tint in shadows and magenta in highlights or (more likely) that's a result of just not enough discreet steps.

cheers
Paul

[Brian Williams]

3 hours ago, rawshooter said:

Unfortunately yes.

Mine doesn’t forget anything. Whenever I turn back on, it’s in the same mode as I left it, with all video and still settings as they were before turning off. 
Now if I’m using one of the custom modes (which I do, C1 for stills and C2 for video) that does reset itself each time you go back to it, but that’s just the nature of custom modes in all cameras.

[Lars Steenhoff]

44 minutes ago, paulinventome said:

As an aside, i did a quick bit of comparison with an SSD plugged in now. I think it was pretty expected:

12 bit DNGs, the best quality both in Shadows and Highlights - the shadows are also not biased green
10 bit DNGs, these are identical to 8 bit in the shadows, band for band. However the highlights roll off with more data in so they match the 12 bit and work with highlight reconstruction better
8 bit same shadow performance as the 10 bit and the highlights reach the same overall range but the values in the highlights are less tonal and this yields magenta tints. So highlight reconstruction not so quite as good as the others.

In all cases i found it near impossible to 'break' the midtones

I wonder whether that linearisation curve in the 8 bit is a little off with respect to each of the channels, with green tint in shadows and magenta in highlights or (more likely) that's a result of just not enough discreet steps.

cheers
Paul

If only sigma could allow 25 fps at 12 bit I would be really happy

[mercer]

On 12/3/2019 at 3:54 AM, paulinventome said:

IMHO most likely crappy debayering combined with some serious 'clarity' adjustments in that last shot.

Whilst i'm all for destruction testing images i also have to balance with i don't shoot like that. I'm more interested in the things that do show all the time, tonality and range. 

If you ever get to push lightroom around on high contrast edges check some really odd debayering artefacts - and this is supposed to be pro-stills level.

The beauty of debayering is the ability to choose different algorithms depending on scene - of course no one actually offers that...

cheers
Paul

If I ever get an FP, I'd use the dng files the same way I use MLVFS with my 5D3... use Resolve for a very rough cut, color corrector and maybe a CST before transcoding to ProRes for FCPX.

It's an $1800 camera, we cant expect miracles here... in fact... for $1800 it's pretty impressive.

With that said... what's keeping me from pulling the trigger on the FP is that the FP seems to have less DR than my 5D3.

Has anyone tested the internal 1080p raw... that's 12 bit, right?

[Lars Steenhoff]

5d Mk3 is 14 bit, and sigma 12bit at 24 fps UHD or 1080.

But that's just theoretical numbers, the sensor also matters, and I know for sure that the 5d has much more shadow noise.

In determining dynamic range, noise is also a factor.  that said I would love to have a comparison with the same lens with the 5d and the sigma in raw.

I cannot make it as I sold my canon.