Just got a new camera - Sony F35

I'd love to some 'motion' tests, people or things moving fast in front of the camera - it'd be great to get a sense for the motion aesthetic of global shutter versus rolling shutter
 
I'm all for a low-light test out in the streets! A landscape resolution test would also be great, in addition to a test of DR in a "real life" scenario. : )
 
Actually i have footage from those tests you are talking becuse just 2 weeks ago I gave a digital cinema workshop and we broight a few cameras:
F35
C300
Scarlet
1dc
5dmk3
D800
A99

Now the problem comparing all these is that some of them are so different in terms of workflow and color spaces that if you really want to compare them you need to spend quite some time in the grading room with a proper colorist. And time means money. Slog and cinestyle cameras require a lut to properly display and grade the footage, but others do not or do not have lits developed yet, so just be careful when grading log footage, dont just go and bring down the blacks like crazy to get some contrsst back in the image.

I'll try to post some tests.
 
Last edited:
^ I'd love to see DR tests with all of those! I've seen the 1DC high iso DR test you posted, do you have that for all those cameras?
 
macgregor said:
Actually i have footage from those tests you are talking becuse just 2 weeks ago I gave a digital cinema workshop and we broight a few cameras:
F35
C300
Scarlet
1dc
5dmk3
D800
A99

Now the problem comparing all these is that some of them are so different in terms of workflow and color spaces that if you really want to compare them you need to spend quite some time in the grading room with a proper colorist. And time means money. Slog and cinestyle cameras require a lut to properly display and grade the footage, but others do not or do not have lits developed yet, so just be careful when grading log footage, dont just go and bring down the blacks like crazy to get some contrsst back in the image.
Click to expand...

Hence why I think apples to oranges tests/comparisons are pointless. A 1DC, 5D, D800 or A99 have nothing to do with a F35. They are made for totally different applications (even if some insist in using them for the same thing) with very different feature sets.

Now comparing a F35, F3 and even Epic makes more sense. All are S35 cinema cameras, S-log on the first two, RAW on the last, which is close enough.
 
Last edited:
Adamsenoj said:
Hence why I think apples to oranges tests/comparisons are pointless. A 1DC, 5D, D800 or A99 have nothing to do with a F35. They are made for totally different applications (even if some insist in using them for the same thing) with very different feature sets.

Now comparing a F35, F3 and even Epic makes more sense. All are S35 cinema cameras, S-log on the first two, RAW on the last, which is close enough.
Click to expand...


Well, in my oppinion you can compare a porsche with a hyundai, they're both cars. Just don't be surprised if the porsche is better at everything except the price. Same with cameras. I can't stand for example blackmagic colours, it shows lots of chroma aliasing, and the sensor is small, but why not comparing it against a monster such as the F35?

Here are the dynamic range results from that test:

DR2.jpg



And here a sample of how each stop was captured (F35 shown here):

f35%20%282%29.jpg


Notice that is recorded DR, not usable DR. The lower end DSLRs for example, record in highly compressed and low color depth codecs. Therefore the last stops get very few bits allocated and posterization occurs. Also noise was not contemplated in this test, and real world shadow detail could be less than in the chart.

We kept most cameras at the official native ISO (F35, 1DC, C300) while others had to be set at a similar ISO setting to mantain the same lighting conditions. Yes, the 5D will be less noisy at multiples of 160, but at the cost of highlight detail. HTP was turned off for this test.
 
Great, Miguel! Did you also do a test of low-light performance? That's actually what I'm wondering the most about regarding the F35, since I've never seen any real low-light footage from it. They tend to light quite thoroughly in Hollywood, for some reason : P
 
Well, I don't want to be rude, but what do people understand for low light?
Underexposed footage shot in a street at night with sodium-vapour street lights?
A low light carefully lit scene with low power lamps but proper exposure?
 
mmm... interesting results, thanks a lot macgregor!

The D800<5D3 doesn't match my tests, but I carried them out at ISO 100 and this can change things a lot. Very cool to see how all cameras were relatively similar in the shadows and it's the highlights that set the best apart from the rest. Completely different from the RAW-stills world.

Your 1DC result matches what Shane Hurlbut said: completely different camera to light for with respect to the DSLRs: lots more headroom in the highlights, but you'll need more fill light.

Any reason for the Scarlet not to be there? Any death threats or something?
 
Hehe, I get your point. Underexposed would be the right word. Just imagine a situation where budget or time restrictions force you to expose darker than you'd want. It could be a night-time vista, for example, or just a fugly city street.
Of course we always aim to light things correctly, but it's still nice to know the limits of our cameras.
 
I didn't know macgregor already did this testing. Although I want to specifically see how the F35 runs with the SLOG F3 and the Epic. So, my thoughts exactly.

Now comparing a F35, F3 and even Epic makes more sense. All are S35 cinema cameras, S-log on the first two, RAW on the last, which is close enough.
Click to expand...
 
Last edited:
Samuel H said:
mmm... interesting results, thanks a lot macgregor!

The D800<5D3 doesn't match my tests, but I carried them out at ISO 100 and this can change things a lot.
Click to expand...

It is a general comparison. But practical in some sense. In the Full Frame 35mm format, it is not uncommon to raise the ISO for getting smaller aperture settings.

Though, some don't even rate the raw image on the mk3 to 12. So interesting to see the video footage getting 12 here. Oh, cinestyle, you are the Air Jordans of canon video.
 
LeavingTheCandy said:
I didn't know macgregor already did this testing. Although I want to specifically see how the F35 runs with the SLOG F3 and the Epic. So, my thoughts exactly.
Click to expand...
Much more interesting comparison, as these cameras are all some what in the same league, whether by price, specs, or performance. It has been a long time since there has been a comparison that doesn't end with "oh man, that $800 camera sure is close!" Which totally negates the purpose of critical camera tests, but makes for good internet entertainment.
 
Much more interesting comparison, as these cameras are all some what in the same league, whether by price, specs, or performance. It has been a long time since there has been a comparison that doesn't end with "oh man, that $800 camera sure is close!" Which totally negates the purpose of critical camera tests, but makes for good internet entertainment.
Click to expand...

Totally agreed. I'd like to have an Alexa and/or F65 to put into the mix too. (too bad I'm not in that tax bracket!) :laugh:
 
macgregor said:
Now the problem comparing all these is that some of them are so different in terms of workflow and color spaces that if you really want to compare them you need to spend quite some time in the grading room with a proper colorist.
Click to expand...

I am happy just to see what the raw output from the cameras look like. This is enough to give me an idea of what I need to do to chase the look I am after.
 
Alright, I think it's time to start a "Show us your F35 rig!" thread... not something I ever thought I'd see :p
 
macgregor said:
Well, I don't want to be rude, but what do people understand for low light?
Underexposed footage shot in a street at night with sodium-vapour street lights?
A low light carefully lit scene with low power lamps but proper exposure?
Click to expand...

Well most of the current crop of digital sensors can handle a properly exposed shot at 1600 ISO and upwards pretty well. Where some of them fall down is in underexposed shots or in the shadows. The problem I have with the 5D MKIII is the fixed pattern noise which doesn't appear to be an issue with cameras like the C300, FS100, and even the GH3 and Nex-5n. Even in well exposed shots I see some fixed pattern noise/digital rain in the shadows which becomes very apparent when panning.

For a good low light test I would suggest 1600 and 3200 ISO exposed at -1 -2 and -3 stops.

I've got a feature shoot coming up in a few months and I'm shooting a lot of low-key and the 5D hasn't performed as well as I'd hoped in tests so I'm looking for a better option.
 
You must log in or register to reply here.
Back
Top