Dragon and Sony F55 – Resolution, pt 3 – Conclusion – The pitfalls of methodology when testing cameras…

So… The scandalous conclusion would be:

There is no significant resolution difference between Dragon and F55
And I guess I could get a fair amount of Internet Hype and hits to my site, through a conclusion like that.

But the only conclusion I can draw is that this is inconclusive.
And if anything, this demonstrates the pitfalls of camera-testing. (a little pet-subject of mine)

Now, hang on, before start flaming me!

I am not saying that any intentional things were done wrong here. Actually I was there and subscribed to the way we did it. So let’s roll back.

In the Alexa comparison, we saw that the Alexa could not outresolve the chart at either 1080 or UHD, with the chart we used and the way we shot it.

So the test wasn’t really all that flawed.
Geoff mentioned he did not want to publish these, because of different FOV on the different cameras.
THAT is a good argument, but in this particular comparison, I would not say that the difference in FOV is significant enough, unless we had ended up with a totally unexpected result.

I have heard the criticism that we shot at T22 and not T5,6, where the lens would have been sharper.
That is true.
The reason for still shooting at T22 were that the UltraPrimes are still “sharp enough” (which the Dragon and F55 UHD vs Alexa results confirm), and we lessened any alignment issues through shooting at high T-stops.

As we see from the Alexa example, that would not influence this particular result, as the Dragon and F55 still outresolves the chart at UHD, while the Alexa did not at neither UHD nor 1080.
T22 is hardly the reason for this result.

But therein lays probably the answer to this result:
The chart we used is probably not good for more than 1080/2k resolution-tests.
To sufficiently check resolution, you need a chart that ideally out-resolves all cameras tested, but AT LEAST out-resolves the resolutions we are testing for.
Without that, we really do not learn anything new.

Does this mean that the F55 is “as good” for UHD as the Dragon and “significantly better” than the Alexa?

This particular test does not give any info about that. It just gives us the info which is supplied by the manufacturers:
Alexa shoots at 2,8k, F55 at 4k and Dragon shoots at 6k. Alexa resolves less than the highres cameras at both 1080 and UHD delivery.

This particular test, does not tell us much more about the resolution from the  F55 and the Dragon. It does give us some info about the consequences of different resolution vs the Alexa. And that the ALexa probably resolves less than 1080, even despite its oversampling.
But that’s all. As I discuss here:
Resolution DOES matter under certain circumstances and shooting conditions, but it is still just one aspect of a camerasystem.

So, after all these pages, I come to the same basic conclusion as Geoff:
The CML/UWE tests were not good for saying anything conclusive about resolution in the +UHD cameras, and thus I will leave that part of the equation.
But that would (if correct) also explain why there was not much apparent difference between the 2k and 4k screenings of the material.
If the material is not of a nature where resolution matters, resolution will not mater.

My guess is that a moving image like this, would tell us a lot more about +1080/UHD/4k acquisition and screening, than a lot of the images we shot in Bristol.

(Click for fullrez)


I still choose to publish this, to point out some possible fault-sources when shooting camera-tests.

If you like me to publish tests and have use of what I do, I do accept and appreciate contributions to the work through Paypal

All imagery on gunleik.com is © Gunleik Groven, unless otherwise stated. For TFP shoots, the images belong to me and the model.
It is possible to buy most images for editorial, commercial or stock usage. I’ll even sell you large prints of them….
Contact: gunleikgroven@gmail.com

No comments yet.

Leave a Comment

You must be logged in to post a comment.