Interesting article: "36 Megapixels vs 6×7 Velvia" and 4 x 5

[timparkin]

Heh you beat me to the mark there... I am deeply interested in the technical side of photography to the point where I understand my medium as much as possible. Then I ignore the hell out of it until something changes..

[timparkin]

Well oversampling is definitely the key to good results. That's why I think 50mp is actually really useful in a DSLR. Not to get 50Mp worth of huge images but to create a smooth transition from broad detail to fine resolution. In fact I've taken to blurring the fine detail a bit on the A7R as it makes things look very natural looking to me.

[timparkin]

Thought you might be interested in the full size 'definitive' D800E vs Mamiya 7 scanned on the Heidelberg.. I kept the D800 as raw pixels so you can see the proper data..



Tim

[Christopher Barrett]

Maybe I missed something, but why Velvia? It has maybe the worst shadow detail of any film ever produced. I'm a little biased, though. After 12 years of shooting chromes for a living, I am so glad to be done with it. Nowadays I either shoot digital or neg.

[sanking]

Maybe I missed something, but why Velvia? It has maybe the worst shadow detail of any film ever produced. I'm a little biased, though. After 12 years of shooting chromes for a living, I am so glad to be done with it. Nowadays I either shoot digital or neg.

Personal preference of course. Or might there be some influence from Joe Cornish?

Sandy

[timparkin]

No it's exposed for the highlights which were very bright on the lightbox and very little light in the foreground. Check the scene out and check the exposure on the lightbox.. I've tested the dynamic range of chromes and they don't match the D800E nor Colour neg.. (no surprise there). So you have to choose what is more important in a scene. In this case we exposed for the highlights.. The black areas on the cameras are about -6 stops, possibly even darker in the deep shadow parts..


Tim

[Greg Miller]

Thought you might be interested in the full size 'definitive' D800E vs Mamiya 7 scanned on the Heidelberg.. I kept the D800 as raw pixels so you can see the proper data..

Tim

Obviously the Velvia wins in resolution; not even close. I have a hard time getting past all the grain anymore - it just sucks my eye in (just a matter of taste; others will disagree). And the shadow detail is all blocked up, especially around the camera strap on the left.

[timparkin]

Personal preference of course. Or might there be some influence from Joe Cornish?

Sandy

Well he shoots the IQ280 now so that's a no..

The Velvia was chosen because it holds more detail than Portra. The new Portra's have actually lost sharpness whilst being optimised for scanning.

Here's the figures from Henning Serger.

Velvia 50: 110 - 125 lp/mm
Provia 100F: 120 - 135 lp/mm
Kodak Portra 400 (new): 80 – 100 Lp/mm

That's a substantial difference (although not as much when scanned obviously)..

If I were photographing the scene as a piece of 'art' I would have given it another stop and a third maybe a let the lightbox blow out. But I wanted the lightbox to be exposed OK to compare some of the finest detail in the picture - the transparencies on it.

Also I still prefer the result from chromes for certain subjects so I carry both. Sometimes I even take a shot on both and use the neg to roll off highlights and shadows on my chrome ;-)

Tim

[Ken Lee]

My remark was more generic in regards to how much time and effort by people on this site spend agonizing over technical aspects, almost to the exclusion of the creative side. Even in the image sharing threads, which for the most part I ignore, where it seems the most common remark is "what lens was that" or "nice tones".

Would you like to see more threads which encourage critique ? How else could we rectify things ?

[timparkin]

Well I might decide to use some noise reduction on that if I were to print but that is always a balance against print size ... as mentioned in my other comment the whole shot was exposed for the highlights on the lightbox and to place the printed res chart at about +0.5. This meant that the unlit cameras had shadows at about -6 or less.

I didn't spend too much time making sure they were opened either so.... as it wasn't a dynamic range test. There is some more info in there but not much.

Tim

p.s. Greg - The grain looks a lot less problematic in a print. and that last image that you can see on screen was the equivalent of about a 245" x 368" print!! (presuming a 100dpi monitor)