Crop or cram? Pixel density versus the big view…
Alpha 380 – 14.2 megapixel APS-C results
The Alpha 380 can not be set to anything except exact ISO steps. At first, I used ISO 400, but found even at 1/1600th the files were better (more) exposed than the other two cameras. It seems that the Alpha 380 sensitivity (on manual exposure, so not in any influenced by metering) is indeed a little higher in calibration, as the brighte results suggested when reviewing the camera itself.
The 1/1600th at f/8 was threatening to leave some burned-out white highlights. Check the full size file, though, and compare it with the Alpha 700 at ISO 320. It’s not so very much worse, and you may even prefer it. I have not used any NR of sharpening on this image, and it could no doubt be processed to look better.
The same ISO 400 setting was also used for the Sigma 70-200mm HSM:
Here you can see the potential for losing a vital part of the shot – the ball! The tighter framing at 200mm is not what I would choose to use, especially with the extra resolution of the A380. Pulling back a bit to allow more space for timing variations is much safer with such fast moving subjects. The A380 fires with a tiny bit more apparent lag than the A900 and A700, both of which are very fast and enable precise ‘catch the instant’ timing. With the A380, a little more anticipation is needed.
The next shot is taken with the ISO dropped to 200, but the shutter only changed to 1/1000th (reducing exposure by 1/3rd stop). I have also used post-processing on the raw conversion, eliminating CA, adding sharpening and removing noise with exactly the same settings as used on the Alpha 700. It’s certainly not as clean a result technically but would probably print better:
Here I’ve got the timing of the A380 a bit better under control, along with composition through the viewfinder which is surprisingly dim after using the other two cameras – not to mention very small. The AF focusing seems to be every bit as accurate. An aperture of f/8 may not seem that demanding at this distance, but when you view at 100%, it’s clear that none of the three cameras or the lenses have FF/BF issues.
Now for the final test, the ultimate walkabout travel camera – A380 plus Sigma 18-250mm. I abandoned attempts to set 200mm and went for 250mm – why not! I reckon you might want to see just what the Sigma image looks like on this difficult CA-generating white subject against a dark background, at full stretch, and the reasonable working aperture of f/8…
Once again, I have let the noise level of the Alpha 380 ISO 200 file remain uncorrected – as with the CA. Here my timing is a little late, but I doubt you could measure the error with a stopwatch. I’d have liked the ball in front of the wicket, near that white boundary marker flag.
Here is a processed version – CA removed, noise reduced, white balance fine tuned, capture stage sharpening added:
It’s worth looking at both images just to see the rainbow colour effects present in the top left foliage and small highlights in the unprocessed version – which disappear when NR, CA and sharpening are all applied.
A very interesting comparison……
Three generations of Alpha – ok, well two and a half, because the A100 was the first generation!
The improvements in the image quality are quite interesting to note. The A900 has better quality than the A700, but the A380 is possibly the best for a small resolution image. It would be interesting to see the direct comparison (at 100%) between the three cameras, with the same lens at the same settings. I am sure that the 24Mpx A900 will produce better detail in a print or overall image, because of the relatively smaller pixels, but the A380 may well surprise us all.
Thanks for taking the time to think this one out!
Yes, I think the 14 megapixel sensor is up to the quality of the A100 sensor, probably better. It has a better handling of highlight to shadow range, and more subtle colour. The colour is the biggest difference, some will prefer the A100. I definitely think it could be used in a better body, and the complaints made about high ISO noise are not entirely justified. It can turn in very good results at 400 and 800.
Hi Davie, good article…
what is your opinion about this 14MPx CCD??
Do you think this CCD could be use in a better body??
Do you think this 14MPx CCD can reach the imagen quality of a100 sensor??