A900 crop vs. A700

[sury]

Would some one help me with enabling size tags. I am not able to set/enable them. I can't post large size images anyway. Here are the links to f8 images (original size in flickr) which is NOT the original size from the camera.

With best regards,
Sury

A700/f8
http://farm4.static.flickr.com/3532/377 ... e6e9_o.jpg

A900/f8
http://farm3.static.flickr.com/2591/377 ... 7d8f_o.jpg

[Argonaut]

Harvey said;
You can do a 360 degree panorama for $100? No tripod? No panorama head? Even stitchnig handheld images for a landscape is still a fair amount of effort.

360 panos are not my thing, but I've done them. I prefer something that looks ungimmicked. I've got a panorama head and have used it, but unless you're shooting lots of straight lines at fairly close range, it's not necessary. Electrons are cheap, so I use lots of overlap. I just set the exposure and pan by hand and with a little practice it's not that difficult. It's not even that difficult to do multiple rows, such as a 4x4 array (8000x12000 pixels minus the overlap using my KM5D - suck on that, FF 24MB users). Of course it won't work for weddings, but outdoors - no problem. Then I hand the shots over to PTAssembler and brew some coffee. 99/100 times the automatic setting can figure it out, even in an array.

There are even programs to eliminate moving figures (i.e. ones that appear in more than one shot when they shouldn't) if you don't want to spend too much time in Photoshop.

Getting a bit off topic here, but the point is that a little elbow grease will save you a ton of money and, IMHO, there are some shots you will get that no 35mm equipment will give in one shot. So, I'm looking for a deal on an a700 for my wildlife shooting.

[Argonaut]

For example, this photo is a pan of 4 vertically oriented shots hand-held with a 24mm lens (36mm equivalent) on my KM5D. Even cropped and taking the overlap into account, the FOV (width anyway) is about like a 12mm full frame shot. Technically, the software dealt with the exposure vignetting very well - you can't see any seams in the blue sky - and I haven't found any physical distortion yet. Esthetically, not so great, but that's the photographer's fault.

[01af]

It's the difference which is surprising, between the A900 and A380.

I don't feel it's surprising. The A900's mirror is 2.3× as large. Let's assume the thickness is the same so it's 2.3× as heavy. The A380's maximum frame burst rate is 2.5 fps, so the A900's mirror, at a maximum frame burst rate of 5 fps, is performing twice as many actions per second. The mirror's lower edge has to travel 1.5× the way in half the time so it's moving 3× faster. The kinetic energy of a moving mass is proportional to the amount of mass and to the square of the speed. So the A900's flipping mirror has 2.3 × 3 × 3 = 20.7× more energy than the A380's mirror.

No, I am not surprised.

I have never thought of the A900 mirror action as particularly liable to produce vibration. It seems that it is.

Well ... compared to the A380, I guess it is. Compared to, say, an old Minolta 9000AF, it's not.

Exposures at 1/25th are consistently better than those at 1/50th on the A900/tripod/no mirror lock combo. This indicates that there is a peak of vibration hitting at around 1/50th ...

Yes ... for that combo. Replace the camera, the lens, the tripod, or the tripod head, and you'll get different results. However, the speed range from, say, 1/2 s to 1/60 s generally is particularly prone to catching blur from the mirror slap---here, mirror pre-fire or lock-up is essential. At slower as well as at faster speeds it usually is much less of a problem, and you can often get away without MPF/MLU ... particularly when not using long telephoto lenses or high-magnification macro set-ups. Stiffer tripods resonate at higher frequencies so 1/30 s or 1/60 s is particularly bad. Cheap, light-weight tripods sometimes are better at 1/30 s but unusable at 1/4 s or 1/8 s---not unlike image-stabilised hand-holding.

-- Olaf

[David Kilpatrick]

Year end accounts and had to sent the government several thousand today, Don. I've shot the comparison sets I need, getting them processed and prepared will take a couple of hours. My on-line VAT is now safely finished and the bank has sent the money, which means I don't get penalties. It would be great to have none of that to worry about but there's a lot to fit in...

David

[David Kilpatrick]

Article now in place - marathon (given the SLOW performance of the server due to the level of traffic on the webpages right now...)

http://www.photoclubalpha.com/2009/08/0 ... -big-view/

My conclusion - if you want good shots, Alpha 900 (the extra space round whatever crop you pick will help with most subjects that move); if you want detail, the Alpha 350/380. The 700 is neither one thing nor the other, but gives a very malleable image (low noise, decent size, very smooth image quality) and will only be better at higher ISO settings. I used ISO 320 for these tests. The results at, say, 1600 would put the A700 well ahead.

David

[David Kilpatrick]

All that will do, Don, is prove the A380/350 is the best - unless for example the test is done at 1600 or 3200. On a static subject I find no discernible difference in focus accuracy between the A380, A700 and A900 (except that if the A900 was focusing incorrectly, it could be adjusted).

The page 4 crops could have been made the other way - taking the A380 as the base.

Here we go - really, these are SO close to being the same subject and are all at the same distance, with the same lens (70-200mm Sigma at f/8 and 200mm):

a380-70200-nativesize.jpg (57.82 KiB) Viewed 4340 times

A380 native size 100 per cent clip

a700-70200-380size.jpg (44.85 KiB) Viewed 4340 times

A700 enlarged to match A380 size (full frame would be 14.2 megapixels)

a900-70200-380size.jpg (42.56 KiB) Viewed 4340 times

A900 APS-C section enlarged to A380 size

I will probably just add these now to the end of the article, as they do provide a different view from downsizing the A380 to match the A900, as I had done.

David

[pakodominguez]

Thanks...taking a look now..but even before I take a look I will make a couple of comments.

#1 At the tele end you often times you are already needing to crop with APS-C...not the other way around and you have said as much yourself so the 'framing advantage' would be minimal at best.

#2 You can always back off the zoom on APS-C for gosh sake if you are using a zoom! Not to mention the fact that is often times more difficult to focus with the a900 since the view of distant critical subjects is more limited.

#3 350/380 has the small viewfinder, worse control layout, worse high ISO, and worse frame rate so often times it's not gonna do the trick as well for action and wildlife.

#4 The 350/380 does not focus as well as the a700 or a900 so you need to consider that disadvantage as well.

-Sonolta

you never happy...

The viewfinder pro/cons applies to the A900 over A700 advantages too. You know that, if you are shooting all day long, having a bigger and brighter viewfinder will make your life easier (and allow you to drive back home safely because you can still "see")

:-p

[David Kilpatrick]

I read the article and the conclusion is sorta whack...and FTR a standing batsman, at fixed position, in bright light, in a wide open field, is not really tough to follow and focus action or wildlife.

Errmm.. who said it was? It is slightly more interesting than a shot of a brick wall, it raises other questions which DO matter (like the field of view of the A900 with a given lens offering positive benefits in this type of photography - which includes most athletics, and even includes baseball which I guess is also a standing batsman in an open field...)

Had the subject been tough to focus, the test would have been useless. All the pix were taken pre-focused, carefully, using manual exposure and minimal camera movement other than for framing purposes.

I'll see if I have time to do a more meaningless test of the A700 and A900, but the results are predictable.

Now if you need to test whether the A900's assisted AF module (slightly improved over the A700) gives it sufficient edge in complex follow-focus action photography (which was not the original question) I'll need to find something else and these things only happen at weekends here. The rugger season starts in two weeks with the Kelso 7s, not sure I really want to be into paying admission and putting up with the drunks and rowdies.

For the record, anyone just aiming on program at the cricket would end up with fairly useless pix. The whites in the sun clip very easily. 1/1000th at f/8 at ISO 200 is 1/3rd of a stop LESS exposure than the brightest mid-day midsummer sun on the beach in Britain (1/200th at f/16). Normal midsummer sunny scenes here need 1/250th at f/11 at ISO 200, not 1/250th at f/16. If I shot 'normal' with any of the three cameras the brightest white was un-recoverably clipped in raw (but most so with the A380 as the ISO is under-rated - 200 is really 250 or even more).

David

[pakodominguez]

All that will do, Don, is prove the A380/350 is the best - unless for example the test is done at 1600 or 3200. On a static subject I find no discernible difference in focus accuracy between the A380, A700 and A900 (except that if the A900 was focusing incorrectly, it could be adjusted).

David, with the A380, did you use LV or the viewfinder?
how fast LV react in this sport-like case?

Regards

[David Kilpatrick]

Including the A380 is not comical. It's the only higher resolution sensor made by Sony. You have shot with the A100, I think, and the A380 is well ahead of that. I could survive professionally with the A380, it is capable of turning in very high quality images even if I don't like the size or the viewfinder. The user interface and controls are actually not as crippled as they might seem and I get on fine with it.

David

[pakodominguez]

The fact that you even included the a380 is comical at best!

-Sonolta

I don't think so: the principal is the same: APS format plus more pixels will give you more reach AND resolution. From the beginning the way you conduced this discussion had been rhetoric: we all knew the answer -but this is not the answer you wanna heard.

Regards

[David Kilpatrick]

All that will do, Don, is prove the A380/350 is the best - unless for example the test is done at 1600 or 3200. On a static subject I find no discernible difference in focus accuracy between the A380, A700 and A900 (except that if the A900 was focusing incorrectly, it could be adjusted).

David, with the A380, did you use LV or the viewfinder?
how fast LV react in this sport-like case?

Regards

Viewfinder. The LV would be quite useless - in this sunlight you can hardly see it, you certainly can't see what's happening with focusing, and you only get to compose about 85% of the shot because it shows a crop of an already cropped screen. I only use the LV when it is needed, for peculiar angles or candid shooting. The whole day using the A380 before doing these shots I did not use the LV once. I guess it would be OK with this sort of subject, but for following moving action it is next to useless - it jerks, smears and drags like any video CCTV camera.

David

[robsphotography]

A very thought provoking article, thanks David! The title “Crop or cram? Pixel Density versus the big view …” can be related back to the pixel density (pixels per linear inch) of each of the three cameras as follows:

A900: 4279 pixels per linear inch

A700: 4617 pixels per linear inch

A380: 4963 pixels per linear inch

For example, the article shows that the image size of the “cropped” A900 image is 3960 pixels wide, in comparison with the full sized 24 mp image of 6048 pixels wide. So, as I said in an earlier post, even after substantially cropping the A900 image, the full-sized A700 image size, with the same field of view as the cropped A900 image, is only 7.9% wider at 4272 pixels than the cropped A900 image size of 3960 pixels.

This can also be accounted for by the pixel density of these two cameras, in other words, the A700’s pixel density of 4617 pixels per inch is 7.9% greater than the 4279 pixels per inch of the Sony A900.

Similarly, the full sized image of the Sony A380 is 4592 pixels wide, which is 16% wider than the A900’s cropped image of 3960 pixels width (or 4963 / 4279 also gives 16%).

(Incidentally, David, on page 1 of your article, it shows the width of an A380 image as 4952 pixels wide, but I think this should read 4592 pixels as shown above).

The above measurements show that, the so-called “crop factor gain” of using an APS-C camera instead of a full frame camera, is substantially offset by the fact that the image width of the Sony A900 is so much greater than the two APS-C cameras in this test.

In addition, with a Sony A900 image, because it shows a wider view using the same tele lens as used on an APS-C camera, I find that this gives more flexibility and options when cropping. And you still end up with similar sized images to the full-sized APS-C images, even after heavily cropping a Sony A900 image!

But, you have to bear in mind that the high pixel densities of the A700 and A380 in comparison with that of the A900 are not always regarded as an advantage! This is probably why the pixel density of the Canon 5D Mark II of 3962 pixels per linear inch is regarded as very satisfactory, because the photosites are larger and permit a far better low light performance than the high pixel densities of the A700 and the A380!

For example, have a look at these DxOMark tests, which firmly put the A900 at the top of the Alpha heap when you test for color depth, dynamic range, and low-light ISO.

http://www.dxomark.com/index.php/eng/Dx ... a-rankings

Regards
Rob
http://www.robsphotography.co.nz/Sony-A900.html
Sony A900 examples and links

[Greg Beetham]

Actually I would be interested also in seeing a demonstration of a bird shot of reasonable magnification in the APS-C frame area taken with the A900 with a decent tele lens that most of us could afford (not the latest G's, even though I could probably afford the 70-300 G), the idea is to choose a lens that is reasonable quality-good optics and doesn't cost the earth, the Siggy 70-300 APO would be a good one for example, or a good secondhand KM tele-zoom or even a prime.
What I would like to see is how the results compare after moving the subject off center (into a corner would be good) after focus and then cropping back the photo to APS-C to show how usefull (or not) FF is at using all that extra frame area...
And also for comparison take a similar APS-C shot of a similar size bird from the same distance, if possible (a chook or a bantam would be ok from a decent distance)
Greg

ps. Good post Don BTW...