Tests on F3
 

Any reaction from you guys who have the F3 to the ProVideo coalition tests:

ProVideo Coalition.com: Camera Log by Adam Wilt | Founder | Pro Cameras, HDV Camera, HD Camera, Sony, Panasonic, JVC, RED, Video Camera Reviews

They found lots of aliasing at high frequencies. Also still plenty of Far Red contamination.

re: Tests on F3
 

its accurate however the aliasing is not an issue in the real world. It occurs on +1000 line repeating patters. That is a very difficult situation to find in nature. (I can barely think of any situations. where it would crop up other than shooting fine line ink drawings drawings for a Art Documentary, for example) Unlike a DSLR that begins color aliasing at +300 lines, very easily occurring in the real world. ( combed hair, herring bone jackets, city land scapes, bricks, tiles, straight lines and skin in a medium close up.)

Alistair chapman got a brick wall to alias on an F3, but since hasn't been able to reproduce it. must have been a fluke perfect storm scenario, between 420 codec, optical pattern width, and mix of colors.

re: Tests on F3
 

I agree with Timur, reproducing this kind of moiré patern in a real life scenario is almost impossible. The reviewer(s) failed however to point out more serious issues with the AF100 which have been very well documented by BBC's camera guru, Alan Roberts.

1. The AF100 does not perform particalarly well in HD (ouch!)
2. This camera will not necessarily always deliver short depth of field, large aperture lenses must be used to achieve that. He is referring here to DOF being the same at comparable focal lengths between an AF100 at F2.8 and a 1/2" camera (EX1 or 3) at full aperture (F1.6).

Read the full report here:

http://thebrownings.name/WHP034/pdf/...onic_AF101.pdf

Thierry.

re: Tests on F3
 

Thierry, is there a benchmark for large aperture lenses that will work with the camera (AF100) under most circumstances?
Would old photographic lenses from f2 to f1.4 do the job, or are we talking high speed cine primes or something?
Sorry to go off topic.

re: Tests on F3
 

I think im the only person that actually likes the AF100..... lol.

I think its a great cam. i like its resolution. its soft but sharp, organic even.


This is the ungraded AVCcam footage. I think it looks good.

re: Tests on F3
 

Did you buy an AF100 also Timur?

By the way what do you mean by "organic". What large sensor cam isn't "organic"?
BTW I just bought Wayne's follow focus on your rec. I'm trusting you man!

Need a mattebox that takes vertical 4 x 5.6 grads though - looking like the Petroff.
Just got 28 thru 85 Zeiss ZF's and backing it up with 3 Nikon Zooms and a Tokina -11-16.
I think even an 18-200 Nikon for running around.

If my camera ever shows up I'll be ready. All I need is some work.

Lenny

re: Tests on F3
 

I have not looked much into lenses availability for the AF100 but my understanding is that there are very few fast lenses available, especially on the wide end.

Thierry.

re: Tests on F3
 

AF100 seems a little off topic in this forum, but Alan Robert's statement about a 1.5 stop difference in DOF between 1/2" and the AF100 does not sound right to me. Isn't it closer to 2.5-3 stops?

17.8mm / 6.97mm = 2.5x "crop factor" = >2.5 stop difference

Re: Tests on F3
 

I have seen the F3 aliasing on another shot, some mesh backed chairs. But it only occurred on one chair out of a whole room full when that chair was in pin sharp focus. I'm not particularly concerned about it as it requires a very high frequency repeating pattern that's in crisp focus, unlike my Canon DSLR which will aliase on just about any fixed pattern or texture.

Re: Tests on F3
 

The footage from F3 from Japan looks from bad to ugly in all senses, terrible event, feel really sorry for Japanese people...
pretty bad chroma and highlights clipping, chroma moiré, if we talk about image quality, getting disappointed in the cam. Low light shots still look very good!
(Having AF100 for good light situations and F3 for low light? :)


Has Alan shot any footage, for himself or for the BBC? "Counting pixels" rarely represents real shooting, news, film or television, picture quality from human eye perception view point! Imho footage from AF100 looks more film like!

P.S. Adam Wilt, thank you for the test!

Re: Tests on F3
 

Allard, while being a talented documentarian, has made some exposure and camera setting choices that are unfortunate if you are trying to judge exposure latitude and overall image quality.

The camera is fine.

Re: Tests on F3
 

If you're shooting for news, or fast turn around television, you won't be using setting that make the most of the camera's dynamic range, just those settings that allow you to transmit the pictures without any colour correction or grading.

Re: Tests on F3
 

I would not based F3 image quality jugement on Vimeo videos but here are a few things to take in consideration. Because of his news network (Aljazeera English) requirement, Matt is filming in 50i (interlace) so, you don't see any of the progressive look. In addition, because of the constrain of news gathering and the need to travel light and work quickly, he is using a single lens, a Tamron 18-270mm F3.5/6.3. So shallow DOF and optical quality is quite compromised. I am sure we will see better F3 pics from Matt once he settles down a bit and goes back to production mode. I think his idea was to demonstrate that the F3 is versatile enough to be also used in breaking news. And even at F6.3, those sit downs interviews look pretty nice by news standards....

Thierry.

Re: Tests on F3
 

We do lots of computer corporate stuff out here in San Francisco. I';m wondering if anyone has checked out aliasing and moire with computer screens or things like closeups of computer chips and boards.

By the way - picking my camera up today!!!! YES

Re: Tests on F3
 

In one of the tests Adam mentions they're getting 12 stops of dynamic range from the F3. Considering Sony is claiming that S-Log brings gains of "up to 800% increase in dynamic range" or whatever their language was... doesn't this imply that the F3 will get as much as the RED EPIC/ARRI ALEXA? Even if they're referring to the F3's S-Log increase as compared to REC 709 with that figure, the fact that it's already showing 12 stops without S-Log is impressive, right? Haven't heard many people talking about it and am wondering what the expectations are with regards to DR once the $3300 firmware upgrade is applied.

Re: Tests on F3
 

Sony have said an extra stop in latitude and a 6db increase in noise for S-Log.

I wouldn't be surprised to see the F3 have more useable stops than anything Red has. The only time a Red can come close to the Alexa's dynamic range is in a Red lab. Actual proper testing has always shown another story though.

Alexa-Canon 7D-RED MX latitude comparison

And that's without ARRI raw too.

Re: Tests on F3
 

Hi David. A 6db "increase" in noise? Weird, my Viper works the opposite way. Shooting FilmStream (same as S-long essentially) actually "decreases" the noise.

Re: Tests on F3
 

Yeah, David, I assume you mean "decrease" in noise?

Re: Tests on F3
 

That's what they said, SNR drops from 63db to 57db in S-Log. The extra stop has to come from somewhere I guess.

Re: Tests on F3
 

I wonder what they did in the FS100 for it to drop all the way to 54db? Man I wish I could afford a F3.

Re: Tests on F3
 

someone on here said the AF100 was organic looking... I believe the exact opposite, even the "best" videos from that camera have just been video looking. The F3 on the other hand is a whole different monster... I wouldn't call the AF100 even close to feeling organic in its look.

Re: Tests on F3
 

I bet it could if the following was done:

1-Get all that detail processing turned off. Most AF100 users don't seem to care enough to dig into this, from what I've seen.

2-Turn off noise reduction. Video noise can be fine, if it's fine enough to resemble film grain and hasn't been "reduced" or otherwise processed by DSP trying to get rid of it but never really does.

A soft, noisy camera can look great if it's done right. I think the key is just to let it be soft and noisy, not try to sharpen it up by adding white halos around everything and reduce noise until it's 10 pixel wide dancing colored splotches.

Re: Tests on F3
 

I see what he is saying. A good field review over at PVC detailed its issues with highlights and this person had to blur the harsh clipping in post which blurred a lot around it as well. Apparently, with many Panasonic cameras, the chroma clipping begins too early and cannot be altered without desaturating the scene profile. For example, a car's taillights were red when brakes were applied but the AF100 showed a mix of red and orange rather than all red. The author compared it to the Auto Knee in EX-cams.

His term for this issue is "saturated-highlight handling problem".

His conclusion was that the AF100 is great with controlled lighting but not good in all other situations.

Btw, has anyone seen the great video by Andy Shipsides at Abel where he tests the DR and shows how the Cine gammas affect the image by using a waveform? AbelCine Tests the Sony F3 Dynamic Range | CineTechnica

F3 has 12+ stops and almost 900 lines of res while the AF100 has 10 stops and ~680 lines of res. I think the AF100's low res is mainly why some people call it organic. I think it was Andy (in another video) who said that S-Log will increase the DR to 13 stops while losing 3 or 6db of sensitivity in addition to much greater detail in the deep shadows and highlights.

Re: Tests on F3
 

Beautifully shot with empathy, however I don't know about anyone else, but I am not that comfortable reading about bad chroma, low light shots and highlight clipping when viewing the content of those clips of the aftermath of the Japanese earthquake and tsunami.

Re: Tests on F3
 

In the provideocoalition test it was closer to 13 stop than 12 (he could see a hint of it in the shadows). So with the S-log it could be closer to 14 stops which would be very near the Alexa. The F3 seems to be of the general family of digital camera that tend to have the most information in the shadows. So generally it is better to underexpose to preserve highlight. This has been shown again in the single camera test at Nab.

Re: Tests on F3
 

Jean, Could you tell more about this test?