Pages: 1 [2]   Go Down

Author Topic: sensor size vs optics  (Read 9610 times)

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
sensor size vs optics
« Reply #20 on: February 02, 2006, 04:45:18 pm »

Quote
I put my Nikon 24mm on the film camera and the Nikon 12-24dx on the digital. I zoomed in to about 16mm which fills up the d2x view finder to be equivalent to the 24mm fixed focus film lens.  The optics are different, the images are not the same, even though they cover exactly the same area. Its is wierd, but obvious - and I don't like the digital.
[a href=\"index.php?act=findpost&pid=57218\"][{POST_SNAPBACK}][/a]
Could you describe what it is that is different, and what you do not like? What f-stops and shutter speeds did you use in each case?

One obvious possibility is that you used the same aperture ratio in each case, which means that 16mm in DX has a smaller effective aperture diameter than 24mm in 35mm format by a factor of 1.5x, or about one stop. Therefore, it gives more depth of field by about one stop. If you want to get back to the lower DOF seen in the 35mm case, open up one stop with the DX lens.

As a side effect, this lower f-stop used in DX format to get "the same composition" (same DOF, same field of view) lets you use half the ISO speed to get the same shutter speed (or twice the shutter speed at the same ISO speed) compared to the f-stop needed to get that composition in 35mm format.


P. S. I suggest you ignore any comparison between different formats based on assuming the use of the same aperture ratio and ISO speed, despite the different focal lengths and DOF involved, unless you never care about getting a particular amount of depth of field.

P. P. S. Sorry, I realize now that this has been said in bit less detail already. To paraphrase, the ultimate solution is to have lenses of large enough aperture diameter to cover one's desires for low DOF and high speed, like 16mm f/2 in DX format to match 24mm f/2.8 in 35mm format. Some opposite extremes:
- Those who use 35mm format f/1.4 primes wide open for very shallow DOF are better of trying to stay with 35mm format for that effect, if they can afford the extra thousands.
- Those who almost always use f/4 and smaller in 35mm format for adequate DOF could shift to about "f/2.8 or smaller in DX" or "f/2 or smaller in 4/3" or "f/5.6 or smaller in medium format" to get very similar results, with the smaller formats being more cost effective.
« Last Edit: February 02, 2006, 05:03:24 pm by BJL »
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
sensor size vs optics
« Reply #21 on: February 02, 2006, 04:52:18 pm »

Saxon, you wonder where it all stops? It doesn't. 35mm film is over-with. Yester-years' technology. Gone where the 33 rpm LPs went. Think of the value of your time. Scanning film and cleaning-up the after-effects is a huge waste of time relative to the ease of getting better quality images from a good DSLR and Photoshop. (I know beacuse I do both, given the stock of legacy colour negs I'm still processing - but once that is complete, no more - finito.) I examined the images on your website. You have many well-composed photographs with great colour balance and luminosity. As far as one can tell from looking at small size images over the internet, I didn't see anything there you couldn't do equally well more efficiently with an appropriate DSLR once you've mastered the work flow. You need to do your own research on this, but from all that film gear you describe, commercial horse-sense suggests to me the sooner you sell it all, the more money you'll get for it - while there remain some people interested in 35mm film; then you can put all the proceeds toward a top-flight full-frame digital set-up such as one of the three full-frame Canons. From there, you have a vast choice of top-quality Canon lenses. As for the Nikon D2X, unless you can get most of your money back, why not keep it as a back-up camera? Many photographers are finding it a really good piece of equipment in numerous applications.
« Last Edit: February 02, 2006, 04:53:50 pm by MarkDS »
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

benInMA

  • Full Member
  • ***
  • Offline Offline
  • Posts: 186
sensor size vs optics
« Reply #22 on: February 02, 2006, 05:28:51 pm »

Also hard to tell from the subject matter on your website but you should not have any trouble with your garden/flower content in terms of corner sharpness or light fall off.

That is largely an exaggerated issue which sounds like it has a lot to do with using the 24-105 f/4L IS lens and maybe a couple other zoom lenses.  There are plenty of lenses that don't have any issues, especially with the prime lenses.   And your work does not appear to be wide open shooting anyway, so even the problem lenses would probably be alright.  Most of us who have shot on film, 1.6x/1.3x, and full frame digital seem to have found that the full frame digital responds just like film in terms of corner issues.  Not a major problem but something to watch for when choosing lenses.

By the same token, looking at your website it is hard to tell what you think is missing in the D2X and the cropped sensor, as it ought to be OK for that stuff too.

But in the end it is perfectly fine to get a FF camera if that feels more familiar to you and feels more like working with the 35mm film camera.   Whatever makes you feel most comfortable will probably give the best results.  And at this point with the 5D the cost is not so much higher as to matter anymore.   If you already have a collection of lenses, the cost premium of buying a 5D instead of a 20D or D200 is about the same as the cost to buy a single high end lens designed for the 20D/D200.   When it cost $6500 extra to get FF it was a hell of a lot more annoying then the $1500-1200 extra it costs today.
« Last Edit: February 02, 2006, 05:29:19 pm by benInMA »
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
sensor size vs optics
« Reply #23 on: February 03, 2006, 02:16:24 am »

Quote
I would say I'm doing it for 3 reasons:

- Shoot in lower light
- Use less depth of field creatively
- Freeze motion

[a href=\"index.php?act=findpost&pid=57315\"][{POST_SNAPBACK}][/a]

Quite right. But all 3 reasons either collectively or separately amount to shallow DoF. No getting away from it.


Quote
Lots of the 5D stuff I've seen is using the crazy abilities of the camera, combining it with IS, and making a dark scene look like it was shot in mid-day sun.  Not my thing.  I'm glad the camera can shoot at those speeds but I am not going to go use it by default.


I agree. I don't see much point in taking a shot by moonlight and making it look like daylight, except as an exercise in what can be done with a low noise camera. But I think you are confusing the low noise obtainable with long exposures on a tripod at ISO 100 and acceptably sharp and low noise handheld shots at ISO 3200. The 5D is good but it's not that  good   . It can't capture detail in dark shadows at ISO 3200 and make them look like daylight   .

Having recently returned from a trip to Nepal, Thailand and Cambodia with my new 5D, I found it quite revealing that I could walk around at night without flash attachment and take candid shots, provided the streets were reasonably well lit. When people don't see a flash going off, they assume that no photo has been taken and tend to relax, even though they might see you pointing a camera at them.

Movement of the subject, however, is a problem which IS cannot solve, so I often took duplicate shots at say f4 and 1/50th and f8 and 1/13th, hoping to get lucky.

The image below is one such shot taken at f8, 1/13th sec and ISO 3200.

[attachment=211:attachment]

The image was converted from RAW using RawShooter Premium. No sharpening was applied (except 'detail extraction' which I guess is a type of sharpening). A modest amount of color noise suppression was used and 'hot pixel/pattern noise suppression'. No processing was done outside of RSP, except resizing and conversion for the web.

For those interested in just how much the image has been degraded through use of ISO 3200, I've provided 3 small crops below. Each of these crops at 100% on my screen is about 5.5x7.5". To see this amount of detail from the same distance to my monitor on a print of the full image, the print would have to be 4 ft wide.

[attachment=212:attachment]
Logged

jani

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1624
    • Øyet
sensor size vs optics
« Reply #24 on: February 03, 2006, 11:03:52 am »

Quote
The image below is one such shot taken at f8, 1/13th sec and ISO 3200.

[attachment=211:attachment]
Ray, was this in the Thamel district in Kathmandu? Did you also like the city (except for the pollution, of course ...)?

The scene is a very nice demonstration of the 5D's capabilities combined with an IS lens.

When I was in Kathmandu in October, handheld shots (or with a monopod) at nighttime were fully feasible with both the 20D and 5D, provided that the ambient lighting was sufficient (as in your photo).

The noise at ISO 1600 and 3200 is very acceptable under these conditions, and I was satisfied with the quality I got myself, and I was pleased with the output of the 5D, too.

(I can't share all images taken with the 5D, since I haven't asked for permission from the photographer in all cases where I didn't take the pictures.)
Logged
Jan

benInMA

  • Full Member
  • ***
  • Offline Offline
  • Posts: 186
sensor size vs optics
« Reply #25 on: February 03, 2006, 12:31:05 pm »

Honestly to me they look like ISO 800 shots done with Fuji NPZ... pretty amazing for ISO 3200.

Your shot looks nice, but I think you know the type I was referring to.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
sensor size vs optics
« Reply #26 on: February 03, 2006, 08:00:00 pm »

Quote
Ray, was this in the Thamel district in Kathmandu? Did you also like the city (except for the pollution, of course ...)?


[a href=\"index.php?act=findpost&pid=57369\"][{POST_SNAPBACK}][/a]

Jani,
Yes, that shot was in Thamel where I spent most of my few days in Kathmandu. I was staying at the Malla Hotel. A right turn out of the hotel took me to the King's Palace and the broad streets of modern Kathmandu. A left turn took me to the chaotic backstreets of Thamel which somehow I found more interesting   .

The noise and pollution didn't disturb me as much as the constant pestering by rickshaw drivers, would-be guides, hawkers and under-nourished children with souvenirs clutched in their hand. But I suspect I was more of a target than other tourists as a result of the expensive looking camera dangling around my neck.

All told, it was a fairly stressful time in Kathmandu and I was glad to get away from the crowds to do a bit of trekking in the mountains. But that was really just swapping one type of stress for another   . At my age and weight, continuous climbing of 800 metres up big steps of granite ( from Landruk to Gandruk) tends to turn my legs into jelly.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
sensor size vs optics
« Reply #27 on: February 03, 2006, 08:28:23 pm »

Quote
(I can't share all images taken with the 5D, since I haven't asked for permission from the photographer in all cases where I didn't take the pictures.)
[a href=\"index.php?act=findpost&pid=57369\"][{POST_SNAPBACK}][/a]

Jani,
Some day I'll probably have the best of my images up on a website, but in general I'm overwhelmed by the sheer quantity of images I haven't yet processed. In the days of film I would be much more careful exposing a frame because pressing that shutter button cost money. With a digital camera, it costs absolutely nothing to press the shutter, except the time to sort through the thousands of images.

I know theoretically it shouldn't take too long to zip through say 4,000 images and select a hundred of the best to work on, but I get easily side-tracked demonstrating some point on LL or comparing RawShooter with ACR etc. One of the joys and advantages of being an amateur I suppose.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
sensor size vs optics
« Reply #28 on: February 03, 2006, 09:45:56 pm »

Quote
P. P. S. Sorry, I realize now that this has been said in bit less detail already. To paraphrase, the ultimate solution is to have lenses of large enough aperture diameter to cover one's desires for low DOF and high speed, like 16mm f/2 in DX format to match 24mm f/2.8 in 35mm format. Some opposite extremes:
- Those who use 35mm format f/1.4 primes wide open for very shallow DOF are better of trying to stay with 35mm format for that effect, if they can afford the extra thousands.
- Those who almost always use f/4 and smaller in 35mm format for adequate DOF could shift to about "f/2.8 or smaller in DX" or "f/2 or smaller in 4/3" or "f/5.6 or smaller in medium format" to get very similar results, with the smaller formats being more cost effective.
[a href=\"index.php?act=findpost&pid=57327\"][{POST_SNAPBACK}][/a]

BJL,
You really should read all the posts before replying   . Many of us now have a pretty good grasp of the DoF issue, thanks in part to your own explanations of the issues as well as Michael's excellent tutorials on the topic.

Perhaps of interest is the fact that my above example at f8 and 1/13th sec exposure does not show significantly greater DoF than another shot of the same scene at f4 and 1/50th. I prefer the f8 shot for compositional reasons but any increase in DoF is difficult to discern. I'm assuming this is a case of image degradation, due to high ISO, increasing apparent DoF.
Logged

jani

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1624
    • Øyet
sensor size vs optics
« Reply #29 on: February 04, 2006, 05:13:56 pm »

Far off in off-topic land, Ray and Jan continue their ruminations.

(I'm not really sure which forum I should select for posting this in as a new thread, it seems to cross the boundaries between this forum and image processing ...)

Quote
Some day I'll probably have the best of my images up on a website, but in general I'm overwhelmed by the sheer quantity of images I haven't yet processed. In the days of film I would be much more careful exposing a frame because pressing that shutter button cost money. With a digital camera, it costs absolutely nothing to press the shutter, except the time to sort through the thousands of images.
Yes, and I find that this is a big enough problem in itself.

To digress further, when I read the article about the LF photographer who tried the P45 back and was all "ooh, aah, now I can take a bazillion pictures and select the best ones instead of having to concentrate a lot to take the best pictures", I was absolutely stunned.

I started out that way when I bought my first dSLR (the 20D in December 2004), and was that way for quite a long time.

But for me it's been useless for getting good pictures in itself; it's been useful only as a learning process.

The best pictures seem to crop up when I make plans, consider what effects I want to get, and take between one and a handful of shots.

This seems to be equally true for my street, architecture, portrait and landscape photography.

Quote
I know theoretically it shouldn't take too long to zip through say 4,000 images and select a hundred of the best to work on, but I get easily side-tracked demonstrating some point on LL or comparing RawShooter with ACR etc. One of the joys and advantages of being an amateur I suppose.
For selecting and processing these images, I think I like both Adobe Bridge and Lightroom. I might like Aperture, too. Lightroom and Aperture would win for selecting the images with the best focus and detail, considering their nifty zoom features.

What annoys me in my hobbyist world, is that I can't seem to find the time to consider my images carefully enough.

I just recently went back to a shoot I had in February last year, because I had a vague recollection about a nice reflected sunset scene that I might be able to do something about. Why hadn't I done anything about it earlier? Well, when I had the time to do something about it, I was out shooting!

So there's this nasty compromise between taking more pictures -- which I love -- and selecting the keepers, winners and portfolio candidates.

I don't expect this compromise to disappear completely if I became a full-time photographer, but I think it's more prominent with us hobbyists.

Well, for those hobbyists who have more than one hobby, anyway. :cool:
Logged
Jan

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
sensor size vs optics
« Reply #30 on: February 04, 2006, 06:56:27 pm »

Quote
But for me it's been useless for getting good pictures in itself; it's been useful only as a learning process.


For me, I find it's this learning process that slows down the sorting and selecting. I've got say 4 shots of a scene; 2 at f11 and 2 at f16; 2 at automatic exposure and 2 a stop underexposed. I know that one of these 4 I want to keep, but which one? (Of course in practice I'm going to keep all 4 because I've already recorded them on CD out in the field and then transferred them all to DVD back home before even beginning to process them.)

Quote
The best pictures seem to crop up when I make plans, consider what effects I want to get, and take between one and a handful of shots.


Or sometimes when you haven't made any plans at all and the scene is just there for one fleeting moment and one hopes that whatever combination of aperture and ISO the camera is set on is sufficient because there's no time to change it.  
Logged

jani

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1624
    • Øyet
sensor size vs optics
« Reply #31 on: February 06, 2006, 05:12:31 pm »

Quote
For me, I find it's this learning process that slows down the sorting and selecting. I've got say 4 shots of a scene; 2 at f11 and 2 at f16; 2 at automatic exposure and 2 a stop underexposed. I know that one of these 4 I want to keep, but which one?
And not to forget 12 other shots of the same scene, but with different angles and FOV.

Quote
(Of course in practice I'm going to keep all 4 because I've already recorded them on CD out in the field and then transferred them all to DVD back home before even beginning to process them.)
Yep.

Quote
Or sometimes when you haven't made any plans at all and the scene is just there for one fleeting moment and one hopes that whatever combination of aperture and ISO the camera is set on is sufficient because there's no time to change it. 
Oh, yes, that too. Though I have a far lower success rate with that.

Well, according to me, anyway. But what do I know? I'm only the photographer and editor, I may be too close and personal.
Logged
Jan

Slough

  • Guest
sensor size vs optics
« Reply #32 on: February 06, 2006, 05:35:50 pm »

Quote
And not to forget 12 other shots of the same scene, but with different angles and FOV.
Yep.
Oh, yes, that too. Though I have a far lower success rate with that.

Well, according to me, anyway. But what do I know? I'm only the photographer and editor, I may be too close and personal.
[a href=\"index.php?act=findpost&pid=57578\"][{POST_SNAPBACK}][/a]

I sometimes take a picture, review it at home, and think "Naaah". Then a few weeks later I realise that I rather like the result. I think it's a case of not achieving what I hoped for, but not realising that the actual alternative outcome is still okay. So I tend to keep images, unless they are obvious junk.  

Leif
Logged
Pages: 1 [2]   Go Up