Question on low light A580 vs A77 (resampled?)?

Specifically for the discussion of the A-mount DSLR range
Forum rules
No more than three images or three external links allowed in any post or reply. Please trim quotations and do not include images in quotes unless essential.
User avatar
anthony_h
Acolyte
Posts: 19
Joined: Tue Aug 25, 2009 10:22 am
Location: Cheshire
Contact:

Question on low light A580 vs A77 (resampled?)?

Unread post by anthony_h »

Hi all,

I progressed from a 6MP KM 5D to a 10MP A100 (after the 5D hit the ground) and suffered the increase in noise and then to the 12MP A700 to improve the noise situation. I take mainly HDR and low light (ISO 3200 or 1600 - astrophotography) shots.

So some years on the 24MP A77 comes out with nominally worse low light performance than last years 16MP A580. I had been waiting to upgrade to the A77 but I am now not sure the best way to go.

One option is the A580 to get great low light performance but have to go back to manual HDR exposure bracketing (I know it has Auto-HDR, but I want to do the tonemapping myself).

The second option is to get the A77, at 24 MP the noise is worse but my question is how does the noise compare when the image is resampled down from 24 MP to 12 MP (after Lightroom RAW conversion, not in camera 12 MP JPEG)?

Of course the A580 is £499 while the A77 is £1000+, but let's ignore that for the moment, likewise video.

Does anyone have any experience with resampling the A77 output and how it might compare?

Cheers,

Ant.
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by David Kilpatrick »

I can't say for astrophotography (Magnar's posts indicate that the A77 is superb) but we use the A580, A55, NEX-5, A900 and A700 as well as the A77. What I'm finding is that for my Alamy submissions, where getting a certain level of sharpness and noise optimised matters, the A77 can not be used with the same complete disregard for all ISO settings up to 1600 as the A580/55/NEX. These, in our workflow, show no significant noise and no loss of sharpness whether the high ISO is for action or for low light.

The A77 shows a very marked, fine, crisp noise level which can not be removed at 1600 without sacrificing some detail. But when reduced to 3600 x 2400 pixels, I can work at up to 3200 on the A77 with results that pass Alamy QC, so far. I'm sure to take it too far some day. I'd say that at this 8+megapixel reduced size, the A77 images look better than A700 reduced, but not better than A55/580; equal perhaps.

Here's a 1600 ISO image (1/80th at f/5.6 wide open on the Sigma 70-300mm OS) which was about 1 stop underexposed in raw - I find this better than using 3200, for the same actual exposure, in this case the longest shutter opening time I could reasonably risk with the subject. It would have been good to be at f/8 or f/11 - but no way.
edinzoolilacbreast2.jpg
edinzoolilacbreast2.jpg (113.39 KiB) Viewed 7178 times
I'm very happy with the fine detail in the reduced size original and to be honest, the full 24 megapixel is pretty good but for this purpose - picture library - it's better to reduce the image size for better detail at 100% and finer noise grain.

David
User avatar
anthony_h
Acolyte
Posts: 19
Joined: Tue Aug 25, 2009 10:22 am
Location: Cheshire
Contact:

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by anthony_h »

Thanks David for your reply, your thoughts were very useful. I think I am leaning towards getting a A580 as my lowlight option and keep the A700 for everyday duty, then look at the A77 again when the prices drop or even the A99!
agorabasta
Viceroy
Posts: 1198
Joined: Wed Oct 22, 2008 7:41 pm

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by agorabasta »

Sorry to interfere, but I wanted to put things into some perspective.

Below is some pretty dark 2-year old shot made with a700, raw NR at Off. Was shot with CWA metering at zero exp bias.
Processed in Lr3.5 and I also turned off the lens profile corrections to avoid its brightening and smoothing in the corners.
First a full image downsized, with and without Lr detail/NR module
Attachments
DSC05511nr.jpg
(237.24 KiB) Downloaded 4117 times
DSC05511nonr.jpg
(232.52 KiB) Downloaded 4117 times
agorabasta
Viceroy
Posts: 1198
Joined: Wed Oct 22, 2008 7:41 pm

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by agorabasta »

Now the dark lower right corner with and without detail/NR, 100% crops, sharpened 'standard for screen' at export.

As you may see, the NR was really mild.

P.S. Forgot to mention - that was ISO3200 :wink:
Attachments
nrcrop.jpg
(173.51 KiB) Downloaded 4117 times
nonrcrop.jpg
(196.45 KiB) Downloaded 4117 times
Mark K
Grand Caliph
Posts: 395
Joined: Thu Aug 14, 2008 4:10 pm

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by Mark K »

A77 shows a lot of details though with high iso noise....

David: do you think a580 performs better shooting in raw) with iso 3200 or above?
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by David Kilpatrick »

I don't shoot 3200 or above, because we need to be able to use the images. That applies to any camera except the Nikon D3/S where I've shot up to 6400 and been able to make good use of the resulting files for a few shots (but not all shots). As you may note, I chose to use -1EV with 1600 on the bird rather than 3200 and normal exposure.

It's hard to tell because the A580 introduces a critical factor, the mirror/shutter action. The lack of mirror/shutter vibration on the A77 means that most of my shots around 1/8-1/100th in low light at 1600 (my usual maximum) are much sharper than similar shots used to be on a A700. A900, and also A580. There's not much difference with fast speeds like 1/250th or with longer exposure like 1 second, but there is an important 'zone' into which a lot of hand held low light high ISO shooting falls - and in this shutter speed range, the A77's vibrationless shooting helps even more than a better high ISO might.

David
agorabasta
Viceroy
Posts: 1198
Joined: Wed Oct 22, 2008 7:41 pm

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by agorabasta »

I have to agree that ISO1600 is a rather hard limit for all APSC. Above that we may get cleanly processed captures only at the expense of resolution.
It may change only with non-Bayer arrays and at much higher sensel densities so that the same colour sensels sit together close than the wavelenght.
alphaomega
Viceroy
Posts: 1196
Joined: Mon Aug 18, 2008 11:20 pm

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by alphaomega »

I conclude (rightly or wrongly) that Sony should have introduced the A65 with the 16Mp new sensor and ditched the A55. That would have given us an uprated A55 with the new EVF and slightly improved sensor. That camera would have had the best of both worlds with latest viewing technology and Sony's best APS-C sensor to date allowing Alamy type quality at 1600 ISO. So, my question to Sony is when are we getting an A60 and NEX-6 both with the NEX-5N sensor????? I think these cameras would sell. I would be a buyer for at least the NEX-6.
User avatar
aramkostanyan
Heirophant
Posts: 97
Joined: Fri Feb 04, 2011 10:00 pm
Location: Zurich, Switzerland

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by aramkostanyan »

I agree with you Viceroy. As I have A55 and want some of the features that A65 does, mostly the EVF or even a better sensor than this one for low light.
Sony A55 + Sigma 17-70 2.8-4 HSM, SAM 55-200
Aram
User avatar
aramkostanyan
Heirophant
Posts: 97
Joined: Fri Feb 04, 2011 10:00 pm
Location: Zurich, Switzerland

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by aramkostanyan »

I agree with you Viceroy. As I have A55 and want some of the features that A65 does, mostly the EVF or even a better sensor than this one for low light.
Sony A55 + Sigma 17-70 2.8-4 HSM, SAM 55-200
Aram
User avatar
Magnar Fjortoft
Initiate
Posts: 44
Joined: Sat Oct 08, 2011 10:41 am

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by Magnar Fjortoft »

I don't really get this with the 16 Mp sensor in favour to the 24 Mp sensor. At high ISO the 24 Mp sensor is as good (or bad) as the 16 Mp sensor, and at lower ISO settings the 24 Mp sensor has an edge with higher resolution. So at high ISO you get the same result with the 24 Mp sensor and at lower ISO settings you get more resolution. This is my experience from comparing fully processed and printed files from the A55 and A77 (Photoshop CS5 and Lightroom).
User avatar
aramkostanyan
Heirophant
Posts: 97
Joined: Fri Feb 04, 2011 10:00 pm
Location: Zurich, Switzerland

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by aramkostanyan »

Mangar - I am happy to hear that. As reading others review of high ISO it seemed that it is worse than 16MP sensor.
Sony A55 + Sigma 17-70 2.8-4 HSM, SAM 55-200
Aram
agorabasta
Viceroy
Posts: 1198
Joined: Wed Oct 22, 2008 7:41 pm

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by agorabasta »

The 16Mp sensor in an a65 body would simply give you the same quality as some a35 does. That a35 is noisier than a55, that's simply because they tried to lower power consumption to avoid overheating issues the a55 had.

So the 16Mp noise would be the same as 24Mp noise. But you lose the 24Mp resolution.

Thus the real way to go is the Nex7 with LA-EA2 SLT adapter.

P.S. Nex with adapter obviously limits your lens choices to Sigma OS, if you need stabilisation.
User avatar
Magnar Fjortoft
Initiate
Posts: 44
Joined: Sat Oct 08, 2011 10:41 am

Re: Question on low light A580 vs A77 (resampled?)?

Unread post by Magnar Fjortoft »

aramkostanyan wrote:Mangar - I am happy to hear that. As reading others review of high ISO it seemed that it is worse than 16MP sensor.
If you just look at pixel per pixel for unprocessed/default raw converter setting images at 100 % or 200 % on a screen, you might see differences. But what you see might be a result of your monitor. Just look at the same samples on another monitor, and you might be surprised by the differences in appearance. After working with digital for many years, my conclution is that you have to compare large format prints (or crops printed on smaller sheets of paper) from fully processed files.

Also, with cameras in the same league, I think other aspects than image quality should be judged since differences in IQ for fully processed images are more or less neglible. More important is: How reliable is the aotofocus, what about ergonomics, how do the viewfinder fit my needs, what about lenses & system, etc. Also, noise can mostly be well controlled with proper exposure.
Post Reply

Who is online

Users browsing this forum: No registered users and 23 guests