Poor resolution on telephoto - is it normal?
 

The Fujinon lens of EX1 is a very sharp one, but with telephoto it loses.

I took a test series with the zoom on the 70, 80, 90 and 99 settings and with iris 4, 5.6, 8 and 11.

I converted the .bmp pics from Clip Browser to .gif and published the f5.6 pics on my own site.

Note the poor resolution on Z90 and Z99.

http://www.lentovision.se/bilder2/Z70f5,6.gif
http://www.lentovision.se/bilder2/Z80f5,6.gif
http://www.lentovision.se/bilder2/Z90f5,6.gif
http://www.lentovision.se/bilder2/Z99f5,6.gif

These results means that I have stopped using zoom90 and zoom99, reducing the max zoom to around 50 mm and the zoom range to around 10X.

My Canon A1 did not show this reduced resolution (at half the price tag).

Is this acceptable?

Should I send it to Sony for a replacement of the lens?

Hej!
I'm in the same position as you. I've just moved up from the XHA1 to EX1 and noticed the same issues with the lens. I sent it back to my dealer and we had a long telephone conversation and I've also talked to other people I know in the trade and I think it is basically that at the top end of the zoom you need to stop down to at least f8 to ensure no soft edges. I noticed that if you go to f11+ the image gets soft too, so the ND filter is my new best friend! Obviously if you still think there is a problem with your camera then send it back to Sony for a proper assesment. Anyone else out there have thoughts on this?....
Regards, Oliver.

Looks like chromatic aberration to me, you have red and green fringing causing the blurring. My EX3 does this too, but not as badly as yours I think, but I haven't shot charts.
F5.6 in on the edge of diffraction limiting your resolution too, F2.8 to F4 seem to be the sharpest.

Never shoot beyond f8 with that lens and I try to stay below that. There is serious bad chromatic aberration when you are at full telephoto shooting contrasty objects. That's what you get with a cheap HD camera - it is the nature of the EX

Actually it's not that the camera is cheap. Generally classed as an excellent machine. You just need to pay $100K extra for an excellent zoom.

Maybe so, but when Canon A1 - for half the price - could give me a decent resolution at max zoom - which is 20x - then I expected at least the same performance with EX1 at 14x.

Perhaps that was to expect too much. Most of you seems to advice me to accept the situation. And resolution in the rest of the zoom range is really good.

I need an EX3 to for wildlife ...

Don't get me wrong i love my Ex-1, but for HD of course its cheap and the trade off is the long end of that lens and the absurd inability to shoot at f8 & above.

I'm not sure that the physics is understood. Diffraction softening is related to the diameter of the aperture, which is related to focal length. If you don't like the fact that diffraction becomes an issue around f/4 then you are actually saying that you want a larger sensor. At least 2/3 inch and preferably like the RED. Of course if you put a 300mm lens on an EX3 then you certainly can use f/8 and higher (in terms of diffraction softening). I get good results at max focal length, so perhaps you should be looking at your back focus or other faults. I have seen a little CA on the edges with bright white objects (but I had to look for it), so make sure you are really checking real world usage. It certainly shows in your tests.

These comments are not "in support of a great camera", just to remind you that every camera has compromises and we have to be realistic in using what we've got. If you look at the cost of excellent HD zooms, then one can judge what ought to be expected in affordable gear. If you're shooting wildlife, then the lenses you'll be putting on your EX3 will each greatly exceed the price of the EX3 body.

EDIT: looking again at the posted gifs, I'm thinking that they really show that the MTF is lower at the long end of the lens.

I'm not techy or do well with viewing 'charts', but I know even at full zoom I get razor sharp images. But focus is obviously absolutely critical at full zoom, it's easier to miss than hit the sweet focus spot. If I'm shooting an animal or my kid at full zoom, at fairly close range, I can tell by the eyelashes when I nail the focus. It's friggin' amazing when nailed, but only so-so when not completely nailed (frustrating because I thought I nailed it in the field), and obviously horrible if I miss critical focus. Sometimes I don't know for sure until I get back and look at it on the big monitor, even though it looks great on the lcd. If possible (and it's not usually practical on a moving subject), I use the expanded focus at full zoom. All my 'tests' are strictly visual though, I'm not shooting charts, so take what I'm sayin' with a shot of Jose Cuervo.

It's been said here - but make your NDs your friend, and always use them well before the silent scream for them occurs in your v'finder.

Diffraction (read: loss of sharpness) starts to occur the moment you start to stop down the lens from maximum aperture. It's not generally apparent for a couple of stops because closing to f/2.8 or f/4 cleans up a lot of the fuzz caused by miss-centered elements and internal flare, but it's there of course.

If you shoot at f/5.6 and smaller apertures you will get softer and softer pictures. This has nothing to do with Sony or Fuji or price, it's all to do with the laws of optics and the way light travels and bends when it meets sharp edges (of which the diaphragm blades are a prime example).

tom.

I here that. I shot a festival in glaring daylight the other day and stopped down instead of engaging the ND filter. When I reviewed the footage back home on a bigger screen it looked almost Standard Def resolution.

Reshooting again in the same conditions today with ND1 and iris open, the sharpness was much better even zoomed right in. Learnt the hard way.

A lot of folk wonder how it is that smaller, domestic cams seem to get away without any apparent switchable NDs. The answer is of course that the smaller the chips, the earlier diffraction losses become apparent and the more ND is needed to void sharpness losses.

Cameras with 1"/5 chips (remember the PDX10?) generally won't let you stop down smaller than f/4.5, and cameras with 1"/6 chips often don't stop down at all once you zoom to telephoto. Although the 'display' readout will show you shot at f/8 (say) the camera will in fact be using max aperture and lots of internally applied ND filtration.

tom.

My problem is not a poor resolution at f8, but at the end of the zoom range. I don´t get better behaviour of the lens at f4.

Serena: what is MTF?

Why don´t a lot of people check the resolution of their lenses as soon as they get them? The smartest guys of this forum could create a manual, including test charts, so that each individual can see if the lens of their expensive (cheap?) EX1 is within what could be expected.

This Fujinon lens has been praised many times in this and other forums, so I did not think of poor tele resolution.

I regret that I did´t do this test long ago. I have got some poor footage because of max zoom, which I then thought was because the motif was out of focus ("how could I miss focus on this clip???").

I can't compare to your Canon,
But I have always felt that the very end of any zoom lens I've ever used was the worst part of the lens. I only just glanced at couple of your shots but it doesn't seem to me the Z99 @ 5.6 was that horrible. The CA was much more noticeable and you lose contrast (to be expected)but it doesn't look like the resolution fell apart terribly.

Maybe I'm too forgiving or didn't look at your charts enough.

MTF: Modulation Transfer Function. It measures the recorded variation of a signal compared to the input. Resolution is ability to distinguish between two adjacent points (so, as an aside, the pixel spacing of a sensor does not represent resolution, since you must record on at least two adjacent pixels to detect a difference between them). In photographic terms, as the spacing between black and white lines diminishes (i.e. the spatial frequency of the line pattern increases) there comes a frequency at which all is grey (the black is recorded as mid grey, as is the white). The modulation response is zero at that frequency and all above that. A good system will hold a high value of MTF and then roll off quickly. MTF in a poor lens will quickly diminish towards zero, although both lenses may achieve the same resolution limit. Google will provide several references, one of which is Modulation Transfer Function - what is it and why does it matter? - photo.net.

Do people check out their lens (and cameras) when they buy? Certainly. That is how I (and others) discovered vignetting in the early production models. And back focus problems. And ND filter problems. And why Sony replaced the lens on my camera.

In doing tests you need to record the conditions of the test; gamma, detail, etc. I've just had another look at mine full tele and although there is detectable CA near the edges this isn't enough to be seen on a real subject. Worst with detail off, really rather good with detail on.

Serena, Unless my lens is a fluke (which I doubt) you will probably see quite serious CA on real world shots at full telephoto where there is a lot of contrast and a bright background . I first saw it with a black clad surfer walking across the frame withy the sky behind him. He had a red line all along one side. The rest of the shot looked great, and the sky was cloudy, so not way over 100% if even that, but it was unusable.
I've also seen it on birds on the beach and other contrasty objects. I've just become wary of full telephoto.

I have not actually done careful tests to sus out when and where it occurs though. Perhaps my iris was closed beyond f8 and that had an effect. I should test it more carefully and if I get the chance I'll report here.

That is CA, not diffraction. Haven't seen anything like that effect (I also shoot yachting events using full focal length). You had a high contrast subject, so it would show any problems.

EDIT: Just to check I took the camera outside and shot this at full focal length, f/3.4. Focussed on the power pole. There is evidence of CA on the aerial on the right; that isn't in the same plane as the power pole.

Incidentally, diffraction limited resolution is a function of aperture diameter (not really about sharp edges) and is quite well explained here: Angular resolution - Wikipedia, the free encyclopedia

For me the way to get rid of the unusable blue fringing with birds against the sky (CA I guess or maybe some other artifact) when using extreme telephoto and cheapo Telephoto adapters was to stop way down, beyond f8 up to f16. That however resulted it very soft images, but at least usable.

You have to be careful about clipping a colour channel, which can lead to very odd effects. Most of the chromatic aberration is contributed by the outside of the lens (off axis optical errors). Closing the aperture reduces the portion failing to bring all colours to a common focus. Similarly errors in focussing can show colour fringes (particularly in 3 sensor cameras).

Serena -the pole on the side was the CA I 'm talking about.
If in focus it would be more noticeable and it is much worse the more off axis you are. If you were following a larger subject ( like that pole) across the frame the CA would follow it across and maybe go away in the center but might be quite noticeable off axis and might be completely unusable if it were the subject of the shot- say a bird or a surfer in a black suit walking across the frame.

Also that is not a high contrast shot. My experience is that if you had a brighter sky ( perhaps an hazy overcast day or a backlit sky) and a subject either black or in shadow the CA would be even more noticeable. I have only seen it in that kind of circumstance but it was very noticeable.
I don't remember ever seeing such bad CA in professional cameras in many previous years, though maybe I just missed it or in SD it was less noticeable.
I still love the camera though.

lenny levy

Yes, I understand what CA is, and the signs of it in an image (as I pointed it out). What I said was that I see it near the edges but not in the central region. That particular frame grab was extracted from a clip where I panned the pole from right to left, and actually employed different gammas and detail on/off. The frame grab shows the power pole and fittings in good detail and in focus without noticeable CA, which contradicts what you've said above. So I do not see what you describe. The subject is quite high contrast (but not backlit) and the image indicates (to me) a lower mtf than I would like. Attached is a frame grab taken today, where the boat fills the frame. There are signs of CA fore and aft, but I don't see that in the majority of the frame. Now I'd like to have the image pin sharp with high MTF all over the frame, but you said "unusable" and I don't find that clip unusable or even questionable (I'd be less happy if it was my Canon 5D).
Discussions of faults or strengths are not a matter for defending a product. I don't make or sell the EX and it doesn't matter to me whether other owners like it or not. Whether or not my experience matches that of other owners is of interest to me and might be of some use to others. I've no doubt that you've accurately described the performance you're getting. I suggest getting the lens checked.

Serena,
I'll have to check my camera more carefully.
The images I describe were not the results of tests but just some shots I took that surprised me. I'll check more carefully in the next few weeks and if I can find that shot I 'll post it.

I think it may require higher contrast and that would be a good thing as it would not be that common , but perhaps there is another explanation.

Are you used to seeing the off axis CA you've seen on the EX-1 on other professional cameras ( not baby cameras but professional cameras that I am telling clients my EX-1 can compete with.)
At least an HVX200 and higher quality. ( don't expect it to beat an F900 .)

Here's something to think about: http://www.biox.kth.se/kjellinternet/Pinhole.pdf

Thanks for the link Serena. I've had a lot of fun with my DSLR. Simply take off the lens and cover the mirror-box bayonet with some aluminium kitchen foil. Use a needle to pin-prick the centre of the foil to make your pinhole, and leave the camera exposing.

Easy and quick to enlarge the pinhole for another attempt, because finding the right size is trial and error. The bigger the chip in your DSLR the more tolerant it seems to be of pinhole size.

tom.

If you are not happy with telephoto resolution you could always use the Adaptimax on an EX3 with Nikon lenses. It gives edge to edge sharpness and way better telephoto. A 200mm Nikon would give 1080mm and a 300mm Nikon would give the equivelent of 1620mm!!! You can see some footage here.
Sony PMW EX3 Nikon Adaptor.