A850 vs. A580

[jfirneno]

I realize that this is a highly specific question but it isn't a random request. I'd like to get a second Sony camera to to use as a low light alternative to my A850. So what I'd like to quantify is how the High ISO (1600 to6400) raw files for each camera compare in low light conditions. I've seen some comparisons of jpegs and some high ISO raw files under normal lighting conditions but I couldn't see very much difference (possibly a one stop advantage for the A580). I'd love to hear from someone who has both cameras and has done some comparisons of this type.

Thanks for your attention
John

 

[jfirneno]

So DxOMark (see link below) apparently says that A850 is still a better high ISO machine than the new sensor in the a580 (if their testing is authoritative). Of course the D3S puts both of them to shame and even the D700 is better than either of the Sony cameras. So I guess that answers my question. Perhaps the new A77 will change the situation and give me an A-mount to use in low light.

Regards
John


http://www.dxomark.com/index.php/en...)/0/(brand)/Sony/(brand2)/Sony/(brand3)/Nikon

 

[jonoslack]

Hi There
I don't have the A580 or the A850, but I do have both the A900 and the A55 - and also the K5 (which have the same sensor as the A580)

I think the difficulty here is that one is always looking at 100% to check out on high ISO performance, and if you do that, then the A55 has at least 1.1/2 stop, possibly 2 stop advantage . . . .... but of course the resolution is less.

I was rather shocked recently when I had to do two wedding books, one shot with the D700, and another with the A900, I was expecting the files from the A900 (shot on a dismal October afternoon) to be much more difficult to process, but actually I found that the extra resolution more than made up for the worse high ISO.

I'm also fascinated to see what comes next - first the A77, and then, hopefully an A99 with a >30mp sensor.

At the moment I'm spending a lot of time shooting the K5, but I'm hoping for great things from sony over the next year.

 

[jfirneno]

Jono:

I'm very interested in your experience comparing low light high ISO shots for the A900 and A55. I have heard many good things about the new sensor vis-a-vis high ISO but I've never seen direct comparison at 100% size of raw shots (without post-processing) to compare noise. I find that above ISO 800 in low light noise begins to creep in. Does the A55 then resist noise as high as ISO 3200?

Regards
John

 

[peterb]

I've read many reviews now of the SLT A55 and only one the A580 (DxO). The SLT A55 reviews seem to say that the A55's DR while good its not nearly as good as the DR on the K5 and D7000, which have the SAME sensor. And they've demonstrated that the noise and NR eat into the detail as well (not sure what happens with RAW files).

Enter DxO. Not surprisingly, according to their testing, the A55, while quite admirable overall, also falls short of the DR and noise suppression at higher ISO's of the D7000 and K5. But, also according to DxO, interestingly, the A580 is right up there with the Nikon and the Pentax. And even though the other two cameras offer features and construction that set them apart, for the most part, with identical sensors (albeit with different image processing algorithms) according to DxO the images they produce in terms of noise and resolution are fairly similar.

I've got to assume that Sony for the most part in two different boxes for their two models, one with a conventional mirror system and one with the new fixed, translucent one, used identical 16 Mp sensors and, I also assume, identical image processing units.

Could it be that the pellicle mirror system affects DR and other matters (I know there's been some reports of ghosting but that must be a reflection anomaly that would result from having a permanent translucent mirror in the light path)? One third less light IS significant and could account for the discrepancy I'm seeing in these test results.

Any other thoughts on this?

 

[douglasf13]

Sure, the pellicle mirror definitely affects various image qualities. However, the advantages of the pellicle mirror apparently outweigh these IQ disadvantages to many. According to DxO Mark, the A55 still has better IQ than any other APS-C DSLR, outside of the D7000/K5/A580, so it isn't exactly poor, and I'd imagine it is fine for the print sizes that most make.

 

[Jan Brittenson]

DR and noise comes mostly from quantization - the sampling of the analog output of the imaging device. DR is defined as the max output divided by he noise floor, so they're intimately related. Improve noise and you improve DR. The cameras also have different filters, and this affects noise as well because if you see low-contrast texture in the image that isn't passable through the filter you know it's noise and can remove it. Quantization noise depends on many factors, including sampling rate. Faster sampling is noisier. A camera that shoots 24MP at 5fps has the same sampling rate as one shooting 12MP at 10fps. The fact that different cameras or manufacturers use the same imaging device from Sony doesn't really matter a whole lot when everything else differs...

 

[douglasf13]

Agreed, and the K5/D7000/A580 perform relatively similarly in this regard. It's the light loss from the pellicle mirror that starts the A55 at a relative disadvantage.

 

[engel001]

I never had a problem with image noise using my Canon Pellix or 1nRS so I am not sure why Sony is having this problem. If I remember correctly, only 2/3 of the light was used for image capture with the Canons.:ROTFL:

 

[jonoslack]

I never had a problem with image noise using my Canon Pellix or 1nRS so I am not sure why Sony is having this problem. If I remember correctly, only 2/3 of the light was used for image capture with the Canons.

HI There
They aren't 'having a problem' with image noise!
all the four cameras using this sensor (D7000,K5,A55,A580) have class winning DxO figures (FWIW). It's just that the A55 isn't quite as good as the others, I would have thought that the pellicle mirror was the obvious reason for this.

BUT - high ISO and dynamic range are fantastic on the A55

As for the A850 / A580 comparison - I haven't really tried it (although I do have an A55 and an A900), but I think the difference in resolution makes any useful comparison rather meaningless.

For example, we all understand clearly that the Nikon D700 / D3 are much much MUCH better than the Sony A900 for high ISO . . . . but my experience in producing wedding books with all three cameras has suggested to me that the extra resolution in the A900 pretty much makes up for the ISO difference for printing purposes.

 

[engel001]

Just having fun, added the emoticon. Someday, I will probably get one of these translucent mirror cameras. It was a truly unique experience with the EOS1n-RS, almost rangefinder-like, to not have the mirror black out the viewfinder during exposure.