Thread-topic: [filmscanners] Re: film and scanning vs digital photography
Tony,
When you say "the image itself has far lower spatial frequencies...", I
guess you mean the image arriving at the film post optics? It is
interesting, the difference between trying to emulate film's grain-clump
quality rather than trying to capture the image itself.
As Oly says, it's not all about pixels. When I went to the Steves-digicam
site, I compared the E-1 5mp image of the hotel to the E-500 8mp image,
comparing the hotel/marina café images on a 23" Apple Cinema display with
both images sized identically. Yes, I know the better comparison is to
print them at say 12x16, since the print is the real test. I just didn't
want to use the ink and paper to do that.
It is hard to say whether the 8mp is better than the other. Of course, 8pm
is not 10mp. I found that some parts of the image looked a bit sharper at
8pm, but other parts of the image looked better in the 5mp shot. Not clear
if it is a depth of field issue, or a lack of identical focusing distance.
The 8 mp had blown highlights, which made them more difficult to compare.
And there were no images shot with the same lens on both cameras. So I
compared another pair of images, and again blown highlights on the 8mp E-500
shot, and other differences like composition and focal length made for a
difficult comparison.
I picked Olympus images for comparison because that's what I have (E-1).
Despite the difficulty in seeing the benefit of more megapixels between
these images, I will probably buy the upcoming sequel to the E-1 when it
comes out, supposedly in the next few months. It will probably be about
10mp. I must say that it is still a question in my mind whether the extra
pixels will be much of an improvement, given all the other factors of image
quality. But I will also move up the their newest highest quality lens (28
- 70mm equiv. f/2.0) if they ever get it out after long drawn out publicity.
That combo may well outperform my E-1.
Berry
On 6/9/07 6:57 AM, "Tony Sleep" <tonysleep@halftone.co.uk> wrote:
> On 09/06/2007 R. Jackson wrote:
>> to fully resolve the grain
>> structure of film takes WAY more resolution than you need to replace
>> it as a capture medium.
>
> Yup. At one time I had 4,000 8,000 and 12,000ppi scans of the same bit of
> film. 8,000 was clearly better than 4,000 (not hugely, but clearly), but
> 12,000 still showed further improvement albeit diminishing returns.
> 12,000ppi recorded the grain topology more accurately.
>
> Now, an information theorist will tell you that's a waste of effort
> because the image itself has far lower spatial frequencies than all those
> pointless wiggly edges of clumps of grain. And they'd be right, except the
> film image *is* the grain rather than what it encodes, and you can see a
> difference with mushy grain that just doesn't look right. But that's the
> difference between photographers and information theorists, taste and
> judgement ;)
>
> None of this matters much if you don't print big enough for it to matter
> or don't care, and I've never longed for more than 4,000ppi personally.
>
> --
> Regards
>
> Tony Sleep
>
>
> ------------------------------------------------------------------------------
> ----------
> Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
> filmscanners'
> or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
> body
----------------------------------------------------------------------------------------
Unsubscribe by mail to listserver@halftone.co.uk, with 'unsubscribe
filmscanners'
or 'unsubscribe filmscanners_digest' (as appropriate) in the message title or
body
