[lol101]

Hi,

I have been using Canon gear (30D + 50f1.4 + 70-200f4IS) alongside my Pentax K10 lately, especially for action shots (mainly my son sporting events) and, because I am a gearhead, I rented a 5D at my local club with the 17-40f4 for the week, you know, just to appreciate the special FF quality...

Yesterday, I tested side by side my K10+12-24f4 with the 5D + 17-40f4 using the same framing of mainly landscapes (forrest).

When I oppened the raws in PS, I was amazed at the level of details and fantastic IQ I got from the 5D!

But then I also developped the K10 pics, did a side by side comparison and... couldn't really find any difference between them, even at 100%, except for the meager pixel count difference.

I did not find more detail in the (excellent) 5D files in these conditions (ISO 100, lenses stopped down between f8 and f11), only when I upped the ISO past ISO 800 would I see the 5D jump ahead significantly.

It was confirmed later on when I looked at my first portrait with the 5D indoor at ISO 1600: excellent sharpness and low noise, the K10 is not able to match this and is a winner in low light.

But frankly, I was prepared to be amazed by the low ISO quality, level of details and high dynamic range of the FF 5D but I couldn't really find one good example of this in the 20 or so shots I took yesterday.

I will test this also with the 70-200f4IS against 50-135SDM and probably also make some comparisons with the 50s f1.4 for portraits and in various light conditions but is it me or are modern APS-C able to match (at least yesterday's) FF quality for landscape shots?

I also thought that if the K20 exceeds high ISO performances of the K10 by at least a stop, I would be interested in seeing a comparison of 5D output to K20 output and would be very surprised if the K20 could not match or exceed it in just about any conditions.


On a side note, it remains true that, generally speaking, the AF and metering (pattern) of the Canons is more reliable than the Pentax's one: some work is still needed on the Pentax side on this.

But I also found Pentax colors to be more accurate (especially reds and blue) and Canon's balance to be a bit too warm (magenta/yellowish) for my tastes but this is largely irrelevant when shooting RAW.

Well, sorry for the rambling but I would really like to know if other people have experienced the same "lack of amazement" as far as landscape pictures abilities of a FF camera: maybe I was expecting too much from my internet readings... or maybe the pixel count is really what matters for this kind of shots after all ...

[k100d]

it was my understanding under well lit conditions and ISO 100, the sensor size doesn't make that big of a difference. even P&S have low noise at ISO 100, e.g. the Optio A30 samples shown in the P&S forum

[regken]

Interesting! The only other comparison I've seen is Rockwell's ( Iknow, I know, he's a jerk) D200, 5D comparison The Full-Frame Advantage
His shows a marked difference but who knows what his agenda is. I'd be very interested in seeing 100% crops of your shots. Any chance you could get your hands on a K20D for the same comparison?

Thanks,
Ken

[lol101]

Interesting! The only other comparison I've seen is Rockwell's ( Iknow, I know, he's a jerk) D200, 5D comparison The Full-Frame Advantage
His shows a marked difference but who knows what his agenda is. I'd be very interested in seeing 100% crops of your shots. Any chance you could get your hands on a K20D for the same comparison?

Thanks,
Ken

One thing to know is that this applies only to raws, jpegs are 100% in favor of the 5D (I find them almost too sharp).

I'll post some 100% crops soon.

Now, for the K20... we have two options:

1/ Ben sends me one of his K20s... I live next to Paris... it's easy... come on!

2/ We start a fund raiser for me to get the K20, purely for comparison purposes of course!

....

[regken]

One thing to know is that this applies only to raws, jpegs are 100% in favor of the 5D (I find them almost too sharp).

I'll post some 100% crops soon.

Now, for the K20... we have two options:

1/ Ben sends me one of his K20s... I live next to Paris... it's easy... come on!

2/ We start a fund raiser for me to get the K20, purely for comparison purposes of course!

....

LOL! Good ones! Now that you bring it up, I bet Rockwell's comparisons are jpegs. The more I think about it the more I am convinced the pixel count is probably more important than the sensor size at low ISO settings, up to a certain size print. The question, I think, is what is that size print. isteve has indicated the K20D gives him a little more latitude when making large prints than his K10D. 24 X 16 inches is as large as I print and am starting to think you would not be able to see any difference between an APS-C and a FF sensor as long as the pixel count are equal.

Ken

[Mallee Boy]

Interesting observations, look forward to the next installment.

However, may I , with respect, also make an observation before this thread degenirates into a shouting match;
that observation is, after a quick search at the cheapest on-line buying source here in Australia (Digital Camera Warehouse):

Canon 5D (body only) : $3299 aud

Pentax K20D (body only): $1525aud

....I would want something better for the extra $$$$$ thankyou very much!! otherwise buying a C 5D would be a waste of money. wouldn't it?

Both are fine cameras and will appeal to many people for different reasons.

Cheers
Grant

[Ivan Glisin]

Well, sorry for the rambling but I would really like to know if other people have experienced the same "lack of amazement" as far as landscape pictures abilities of a FF camera: maybe I was expecting too much from my internet readings... or maybe the pixel count is really what matters for this kind of shots after all ...

I think FF abilities for landscape photography have nothing to do with sharpness. As you have properly observed, if you achieve the same FOV there is no real difference. Resolving power (sharpness) remains the same assuming identical sensor resolution and FOV regardless of the sensor size. So pixel count is what really matters, not sensor size. (Sensor resolution vs. sensor size vs. pixel size = impact on noise, is not a real issue: I can hardly imagine anyone shooting landscapes on a tripod at ISO3200. At ISO100 all current APS-C and larger sensors are very good anyway, and you have correctly observed that as well.)

The main reason for associating FF with wide-angle landscape photography is that if you want wide angle you can use wide angle lenses to it's full potential, while on cropped sensor they become less than wide so you need to buy "digital" wide lenses for the same FOV, but often bigger, heavier and costing more. Given the same registration distance for FF and APS-C ultra wide angle lenses are still retro focus lenses with design challenges that are getting more and more complicated with shorter focal lengths regardless of the sensor size (elements can be smaller for smaller image circle but complexity keeps increasing with shorter FLs regardless of the image circle).

For example, provided that Pentax has FF digital body using an A 15/3.5 would give you a really super wide angle view. With APS-C you will need DA 10/3.5 (rectilinear lens, DA 10-17 does not count here!) which simply does not exist. (Not from Pentax at the moment.)

Or, you can get moderately priced A 24/2.8 for excellent wide angle shooting on FF, while for the same effect you will need again non-existing DA 16/2.8, or slower, bigger, heavier, and up until recently more expensive DA 16-45/4 (now cheaper) and pay premium for 67mm polarizing filter as opposed to 52mm filter. Or go with a bit wider but much more expensive DA 14/2.8. And all that just to get similar result you can get with A 24/2.8 on a FF body.

So I believe FF is popular for that main reason: small, light, cheap(er) and very good 24mm or 28mm lenses again work just fine as good wide angle landscape lenses.

I would love to try my M 28/3.5 on a FF digital body! I doubt DA 18-55 @ 18/3.5 would match the quality (as the closest FOV equivalent on APS-C).

---

Last edited by Ivan Glisin; 04-04-2008 at 03:31 PM.

[Shashinki]

well... 5d is a pretty old camera tho.. released in 2005... aps-c has to catch up at some point.

when will they release the new 5d? the nikon D3 already stole canon's new 1d mkIII and 5d customers so they will have to do something asap :P

(on a side note: too bad that the only canon lenses that have high enough res for a FF sensor with more Mpix than the current 5d is their 50mm f1.2, their 85mm f1.2 and the 70-200 f2.8 without IS or you might as well use the old one)

[RiceHigh]

Well, sorry for the rambling but I would really like to know if other people have experienced the same "lack of amazement" as far as landscape pictures abilities of a FF camera: maybe I was expecting too much from my internet readings... or maybe the pixel count is really what matters for this kind of shots after all ...

I would share my experience to explore the full potential of the 5D:-

1. Use Neutral mode of the Picture Style (PS) which has the best colour accuracy and a normal tone curve, then more DR and more color accuracy can be seen;

2. But since Neutral has decreased colour saturation and no sharpening at all, add +1 to the saturation (to bring back normal colors) and +3 for the "sharpness" (as most other PSes defaulted);

3. Since ALL PSes of the 5D has a higher contrast (if you are to calibrate the camera with a grey scale chart), I set -2 for the contrast for even more DR and get more natural images which my eyes can see;

4. The full potential of a FF DSLR is in the Depth of image which it can give you, no matter for landscape using wide and for portrait using mid-tele;

5. The weakest spot of 5D is soft corner and obvious vignetting when using super wide and at wider apertures. But stopping down will help somehow. Normally, we won't use full opened for ultra wide, unless handheld under very dim condition, I think;

6. Other camera performance aspects should be concerned: 5D has a much faster AF system under virtually any conditions, its AF accuracy is not affected even under very dim tungsten light, it has a much shorter system time lag, it tracks randomly and/or fast moving objects much better and etc.

7. As you say, the metering and exposure accuracy and consistency is far better for the 5D, which I think is very very important and crucial.

Well, more questions are welcome and I am glad that I could help :-)

[nostatic]

aren't those settings for jpg? They don't affect raw images, do they?

I don't think you can explore the full potential of any digital camera if you're taking lossy files out of it...

[RiceHigh]

aren't those settings for jpg? They don't affect raw images, do they?

I don't think you can explore the full potential of any digital camera if you're taking lossy files out of it...

I think the jpeg format for the loss of image data should be the last thing to consider if the camera has got things right out of the box for the first time. Even if you shoot RAW, what format will you convert for an image, usually?

Also, just don't think RAW is completely "raw", which is non-existent on this planet - they just have more bits afterall.

[nostatic]

I think the jpeg format for the loss of image data should be the last thing to consider if the camera has got things right out of the box for the first time. Even if you shoot RAW, what format will you convert for an image, usually?

Also, just don't think RAW is completely "raw", which is non-existent on this planet - they just have more bits afterall.

if you're taking jpg images out of the camera, then you aren't really seeing what the camera is capable of. It isn't a point and shoot, but maybe that's the way you use it?

I would *never* judge a dslr camera based on the jpg output. The point of using a dslr is that it gives the user control over the final image. Work that I print never sees jpg.

[RiceHigh]

if you're taking jpg images out of the camera, then you aren't really seeing what the camera is capable of. It isn't a point and shoot, but maybe that's the way you use it?

I would *never* judge a dslr camera based on the jpg output. The point of using a dslr is that it gives the user control over the final image. Work that I print never sees jpg.

Yes, I shoot 95% of the time in jpeg mode with my 5D whilst with my K100D I shoot nearly 100% of the time in RAW mode. So, according to your proposition, my 5D must be a P&S and my K100D must be a pro DSLR! ;-D

[jeffkrol]

I think the jpeg format for the loss of image data should be the last thing to consider if the camera has got things right out of the box for the first time. Even if you shoot RAW, what format will you convert for an image, usually?

Also, just don't think RAW is completely "raw", which is non-existent on this planet - they just have more bits afterall.

More bits the better. There can be little arguement between 256 colors vs 4096.
Who in their right mind even uses a monitor at 8 bit color? When it comes to photography for more than just snapshots I really see no need in ever crippling your pallete shooting jpg. Not to even mention the compression aspect...
Yes, jpg is "fine" but why?????????
Maybe just me but I never ever will understand the jpg vs raw arguement. Converts your image to the equivalent of a polaroid (instant film) as compared to a true negative/slide. If your camera shoots great jpgs, even more reason to shoot in RAW. It is possible that your images could be ever so slightly better..... and converting to 16bit Tiffs w/ a jpg "preview" or "for the web" gives you everything. Not just a shadow of the data...
Clarkvision: Digital Camera Raw versus Jpeg Conversion Losses
To be honest the above link has enough info to argue in both directions. Bottom line though is why? If the image is just fine, convert it to jpg LATER, when you have time to critique it more and then if you so choose, delete the RAW (I'd wager 70% would still save it, mostly, just because.. ). Storage is dirt cheap and multiple cards (rather than have 100's of files on one) is a bit more insurance against complete loss....

[lol101]

I would share my experience to explore the full potential of the 5D:-

1. Use Neutral mode of the Picture Style (PS) which has the best colour accuracy and a normal tone curve, then more DR and more color accuracy can be seen;

2. But since Neutral has decreased colour saturation and no sharpening at all, add +1 to the saturation (to bring back normal colors) and +3 for the "sharpness" (as most other PSes defaulted);

3. Since ALL PSes of the 5D has a higher contrast (if you are to calibrate the camera with a grey scale chart), I set -2 for the contrast for even more DR and get more natural images which my eyes can see;

4. The full potential of a FF DSLR is in the Depth of image which it can give you, no matter for landscape using wide and for portrait using mid-tele;

5. The weakest spot of 5D is soft corner and obvious vignetting when using super wide and at wider apertures. But stopping down will help somehow. Normally, we won't use full opened for ultra wide, unless handheld under very dim condition, I think;

6. Other camera performance aspects should be concerned: 5D has a much faster AF system under virtually any conditions, its AF accuracy is not affected even under very dim tungsten light, it has a much shorter system time lag, it tracks randomly and/or fast moving objects much better and etc.

7. As you say, the metering and exposure accuracy and consistency is far better for the 5D, which I think is very very important and crucial.

Well, more questions are welcome and I am glad that I could help :-)

Thanks for the reply.

I always shoot raw with DSLRs (as I said, jpeg output with the 5D is better than K10 but since I don't use jpegs, I just took one or two test shots).

It's true than 5D is faster at AF in just about any conditions but I am actually wondering if it really matters since I never have to "wait" for the AF on any of those cameras... I'll explain what I mean in pictures later today when I'll finish PP'ing my last batch (don't wait for any quality shots though... it's been raining all day... )