A57 or A65 for lowlight wildlife?

Started Oct 28, 2012 | Questions
philbot Contributing Member • Posts: 846
Re: A57 or A65 for lowlight wildlife?
2

Surely, the slight NR applied to the a65 should just destroy all the underlying information, and you'll be able to show where the 'pure' A57 image holds the advantage with ease, assuming you are correct that is.

If you want to progress your argument, why not show where the NR'd A65 image is clearly suffering in its lack of image information at ISO3200 compared to the 'pure' A57 image. That's all we ask.

It's no good just pointing out the obvious difference in visible noise OOC in light of what dpreview themselves say about ACR, or are you really saying that noise is more important then the information within it?

sensibill
sensibill Veteran Member • Posts: 5,401
Re: A57 or A65 for lowlight wildlife?
2

philbot wrote:

Surely, the slight NR applied to the a65 should just destroy all the underlying information, and you'll be able to show where the 'pure' A57 image holds the advantage with ease, assuming you are correct that is.

Who cares? The chroma performance of the A65 is worse at high ISO. Whether I can apply NR to an A65 RAW and show detail destruction has nothing to do with the fact that no such noise reduction can occur without a quid pro quo of detail loss, regardless of quantifying its severity.

Or are you trying to say that chroma NR is lossless?

If you want to progress your argument, why not show where the NR'd A65 image is clearly suffering in its lack of image information at ISO3200 compared to the 'pure' A57 image. That's all we ask.

As I have told you before, I am not here to prove a negative and I don't feel my argument needs progressing. Certainly nobody else here has been doing that, so why should I? I think the data speaks for itself.

It's my view that the A65 has quite good image quality in general and there's no need to make a huge deal out of it having slightly noisier high ISO than 16MP models by making a mockery out of a NR process. I don't think post processed files are relevant to this discussion. That there IS (some) more chroma noise at baseline is, however, a fact.

It's no good just pointing out the obvious difference in visible noise OOC in light of what dpreview themselves say about ACR, or are you really saying that noise is more important then the information within it?

Moreover, do you agree with your 24MP compatriot here saying DPR is utterly useless and that given any variance in pipeline processing IC, no such comparisons or baselines exist that are relevant?

You and swede disqualify whole websites based on minor exposure or testing variance if it suits your need to dismiss specific results or images, but then similarly dismiss 'minor differences' in chroma noise as 'easily removed' in post. I suppose the double standard there isn't obvious to you..?

This really is getting ridiculous. And going nowhere. I'm out - you and theswede can compare notes at your leisure.

 sensibill's gear list:sensibill's gear list
Nikon 1 J1 Sony a7 II
philbot Contributing Member • Posts: 846
Re: A57 or A65 for lowlight wildlife?
2

sensibill wrote:

philbot wrote:

Surely, the slight NR applied to the a65 should just destroy all the underlying information, and you'll be able to show where the 'pure' A57 image holds the advantage with ease, assuming you are correct that is.

Who cares? The chroma performance of the A65 is worse at high ISO. Whether I can apply NR to an A65 RAW and show detail destruction has nothing to do with the fact that no such noise reduction can occur without a quid pro quo of detail loss, regardless of quantifying its severity.

Or are you trying to say that chroma NR is lossless?

No, chroma NR isn't lossless, but if I apply it to the A65 and I end up with no less chroma information then the a57, then clearly dpreviews comments on ACR are correct (or the obvious notion of raw preconditioning are correct)..

So please, just show me where this loss is relative to the A57?

This ridiculous belligerent stance of just showing the OOC noise despite even dpreview telling you to expect differences that aren't down to purely sensor performance is raising my eyebrow somewhat, at least explain yourself on this, why must we ignore this factor?

You'd think it was rocket science or something. I've never said to ignore any sites data, just put it into context and don't ignorantly take it at face value..

sensibill
sensibill Veteran Member • Posts: 5,401
Re: A57 or A65 for lowlight wildlife?
2

How about putting this ignorantly at face value: The A65 and A77 produce more noise at high ISO. You can't say post processed images are a valid basis for comparison but that a 1/200 exposure difference of OOC RAW somehow is responsible for chroma performance and negates an entire methodology and the tests that use it.

I did this same test with the Imaging Resource samples - EXACT same results. More baseline chroma noise at equalized resolution. Period. The A65 doesn't have more detail at 16MP and the A57 doesn't 'lose shadow detail'.

You had it right 5 months ago. Sad to say that since then you've deluded yourself somewhat on this issue.



 sensibill's gear list:sensibill's gear list
Nikon 1 J1 Sony a7 II
theswede
theswede Veteran Member • Posts: 4,009
Re: A57 or A65 for lowlight wildlife?
1

I said same ACR, not same internal algorithms. Not that you have presented any data on how A65 ACR conversions differ to account for mystifyingly worse chroma noise performance than A57. It must be a conspiracy.

The chroma noise difference is due to internal camera processing differences.

And thank you for supporting my point; there is no way to know what ACR does internally, and the null hypothesis is that it is not intended for benchmarking but a commercial product for producing results and is therefore not carefully tuned to provide comparable output in its default settings.

How exactly does a sensor array employ noise reduction prior to the imaging pipeline?

You seriously don't know this? Yet you argue about sensor performance?

Go read a few basic articles on sensor design. I'll wait.

So no baseline exists, so you can't say the A65 doesn't produce equal results in any form, and no sensor comparative evaulation can exist. According to your logic.

For purposes of subjective evaluation this is correct.

And we need to know what exactly is going on in order to say Camera A has cleaner high ISO than Camera B why?

Because otherwise we do not know whether Camera A has "cleaner" (a word you have not defined, so it is as of yet meaningless) high ISO than Camera B. The only way to know that is to isolate "cleaner" from all surrounding variables. Otherwise you're looking at garbage.

Per the logic of science. Or as you put it, my logic. Same thing.

Start by defining "cleaner" with specificity so that it can be isolated. Not that anyone cares about anything but actual SNR since everyone who exploits high ISO performance will do PP - it's unavoidable - and then the information content is what matters, not how the information loss has materialized. Most people are unaware of this and look at the wrong things, but that doesn't make them right.

Again, your view is, that either A65 has equal noise performance at high ISO normalized or all such comparative efforts are void in the absence of precisely identical pipelines and associated algorithms both in camera and out. Which is it? Not that it matters.

No. My view is that in all sensor isolating tests the difference between the A65 and the A57 is so small as to be practically unmeasureable; the difference between individual cameras of one model will be larger than the objectively measured differences between the models.

As I have stated numerous times, objective sensor comparison is the only way to compare sensor performance. Subjective OOC evaluation does not compare sensors since there are too many uncontrolled variables. It's not that complicated. Basic science.

The rest of us will suffer with flawed comparisons that show us performance baselines that either exist or not, but if they do, they don't matter, according to you. This will be great news to owners of first gen 12MP Micro 4/3 cameras when they learn that the Olympus OM-D and GH3 have nothing (relevant) on them in terms of IQ. Indeed, image quality as a concept is moot.

If you can't tell the difference between these various cameras in finalized images no matter the shooting circumstances then yes, there is no IQ difference between them. By definition. Not that hard a concept to grasp, I would think.

Since you can you're ignoring my entire point in your little rant. Spectacular performance.

All according to you.

Have fun shooting down strawmen.

Jesper

 theswede's gear list:theswede's gear list
Konica Minolta Maxxum 5D Sony SLT-A37 Sony Alpha NEX-6 Sony a7 Sony a5000 +7 more
theswede
theswede Veteran Member • Posts: 4,009
Re: A57 or A65 for lowlight wildlife?
1

No, they aren't. No, the screen shots do not support your point (which has been that there is no difference).

Considering my point is that it's negligible difference they do support it.

There's a difference, is my point, and that people should make up their own minds, not to have it made for by 24MP shrill contingent shouting down anyone saying the 16MP is cleaner at high ISO.

Nor to have it made for by the 16MP shrill contingent shouting down anyone saying the 24MP performs so close to the 16MP any difference is moot.

Not in the least. DPR can't even keep focus points consistent, much less amount of light entering the camera, so their "comparisons" are a useless pile of junk.

Your opinion might be classified as such.

Fact. Not opinion.

That's not hyperbole. That's how it is. The hyperbole starts after that, when explaining how much time has been wasted arguing over differences which are down to uncontrolled variables and not cameras.

Arrogance.

Indeed. That does not make me wrong.

Hilarious. So we should all have first gen 1" or 4/3 cameras because all sensor performance is irrelevant. Why ever upgrade? Why bother with ANY such tests? Why do you participate at DPR if they're so vile?

It would behoove you to argue against what I state instead of what you wish I would have stated.

And when have I stated that DPR are vile? I would like a direct link, please. That is a pretty serious accusation which I take offense to.

You have not and cannot prove the A57 has less shadow detail at normalized resolution. Because it's a falsehood, like much else you've been posting on this subject for months (years?).

And you cannot prove the A57 does not have less shadow detail at normalized resolution. These are subjective evaluations, not subject to proof.

However, the DxO results clearly show comparable high ISO information between the cameras. Information loss (collectively known as noise, and no, chroma noise is not the sole kind of noise which impacts an image) affects both cameras to a comparative degree. In the A65 it is seen as chroma noise and in the A57 it is seen as blur noise. Both are as devastating for the information content of the image, and both can be remedied by PP to comparative extents.

Such as Phil's. And to say nothing of individual opinions summarily dismissing entire website methodology (DPR) and selectively hand picking only the results from other sites they agree with (like DxO).

Considering you've admitted to starting from the explicit assumption that the 16MP performs better at high ISO than the 24MP while those who examine the issue carefully start from no assumption at all that's pretty rich.

The dismissal of the DPR methodology is because it is objectively (and by DPR admittedly) flawed. It is useless for any pixel peeping or actual performance comparison because it has too many uncontrolled variables. This is not controversial; it is basic science methodology and something kids learn in high school how to do properly.

That you refuse to understand this (and argue about sensor performance without even knowing how sensors work) speaks volumes for your credibility.

Jesper

 theswede's gear list:theswede's gear list
Konica Minolta Maxxum 5D Sony SLT-A37 Sony Alpha NEX-6 Sony a7 Sony a5000 +7 more
theswede
theswede Veteran Member • Posts: 4,009
Re: A57 or A65 for lowlight wildlife?
1

How about putting this ignorantly at face value: The A65 and A77 produce more noise at high ISO.

According to DxO data they're so close that the difference is imperceptible.

You can't say post processed images are a valid basis for comparison but that a 1/200 exposure difference of OOC RAW somehow is responsible for chroma performance and negates an entire methodology and the tests that use it.

Chroma noise is only one kind of noise. You can't compare overall noise by only comparing chroma noise. All you see when looking exclusively at one kind of noise is the decisions made in the in camera noise handling algorithms. You learn nothing at all about total information content in the image.

I did this same test with the Imaging Resource samples - EXACT same results. More baseline chroma noise at equalized resolution. Period. The A65 doesn't have more detail at 16MP and the A57 doesn't 'lose shadow detail'.

And again, all you're looking at is one kind of noise. A completely irrelevant comparison.

Jesper

 theswede's gear list:theswede's gear list
Konica Minolta Maxxum 5D Sony SLT-A37 Sony Alpha NEX-6 Sony a7 Sony a5000 +7 more
Nordstjernen
Nordstjernen Veteran Member • Posts: 6,876
Re: A57 or A65 for lowlight wildlife?

danny006 wrote:

The 16MP sensor has less noise at pixel level, but don't know if you will get better image quality untill ISO1600. Above ISO1600 I'm sure it will perform better but those high ISO you better don't use anyway.

Depends highly on HOW you compare and WHAT you compare.

Pixel by pixel?

or

Senor area?

Pixel by pixel the lower density sensor will look a bit cleaner, but when comparing sensor area there is no practical difference at all. Also a denser sensor gives higher resolution, which is very welcome when shooting at lower ISO. Or do you just shoot at ISO 6400 and higher, and then compare pixel by pixel???

 Nordstjernen's gear list:Nordstjernen's gear list
Sony Alpha a99 Sony a7
Nordstjernen
Nordstjernen Veteran Member • Posts: 6,876
Re: A57 or A65 for lowlight wildlife?

Steve West wrote:

Having an A77, I would recommend the A99 for lowlight wildlife--esp if you are going to crop.

Also note that the A99 has lower pixel density than the a77 so resolution will be lower when the subject is filling the same area on the sensor. This comes into play when shooting wildlife with the same long lens on both cameras, and you need to crop also from the aps-c frame.

That said, the A99 is performing very good at high ISO. I will do some side-by-side ISO 6400 tests with the A77 and A99 later today - both for pixel by pixel, and for total sensor area.

 Nordstjernen's gear list:Nordstjernen's gear list
Sony Alpha a99 Sony a7
Amateur Sony Shooter Veteran Member • Posts: 5,433
My thoughts

A fast 300mm lens is not cheap, even buying third party brand. You need to consider various facts: viewfinder (OVF vs EVF), burst rate, buffer depth, camera high ISO noise, AF performance and configurability, AF micro adjustment feature, overall build quality, after sales support, lens market, etc. The only Sony camera I think is "wildlife worthy" is the A580, or the new full frame A99. Again, depending on what lens you get, you might have to change brand too. In my case, Sigma 120-300/2.8 OS HSM is the single mighty lens I ever need for anything sports/wildlife, but there is no Sony version, I settled with Canon 7D and am generally very satisfied with my choice.

 Amateur Sony Shooter's gear list:Amateur Sony Shooter's gear list
Sony RX100 Sony a77 II Sony 70-200mm F2.8 G Sigma 10-20mm F4-5.6 EX DC HSM Sony 24-70mm F2.8 ZA SSM Carl Zeiss Vario-Sonnar T* +6 more
philbot Contributing Member • Posts: 846
Re: A57 or A65 for lowlight wildlife?
3

sensibill wrote:

How about putting this ignorantly at face value: The A65 and A77 produce more noise at high ISO. You can't say post processed images are a valid basis for comparison but that a 1/200 exposure difference of OOC RAW somehow is responsible for chroma performance and negates an entire methodology and the tests that use it.

Well that does seem ignorant of the RAW Pre-conditioning factor..

I'm not saying it negates anything.. If you approach it logically, and I'll borrow from DPReview again, "so inevitably we are still looking at a balance of noise and noise reduction, rather than pure noise levels", That's two variables, noise and noise reduction (or noise and image information left over).. It's quite obvious and normal to 'fix' or (equalise) one and measure the other.. which is all I'm doing (fix/equalise the noise, and then let people see the difference in image information)..

I did this same test with the Imaging Resource samples - EXACT same results. More baseline chroma noise at equalized resolution. Period. The A65 doesn't have more detail at 16MP and the A57 doesn't 'lose shadow detail'.

That would be lovely if your 'test' wasn't so meaningless.. Remember that it's easy to show RAW Pre-conditioning (I'm classing that as any variability in the signal conditioning from sensor to RAW image conversion), the evidence for this variability is profound and support directly by DPReviews own words..

You had it right 5 months ago. Sad to say that since then you've deluded yourself somewhat on this issue.

Luckily I have the ability to absorb, understand and approach things logically and scientifically, I took a lot to change my ignorant pre-conceived belief, which was based on mis-interpretation of other peoples (dpreviews/imaging-resources) images..



Yeah, lovely.. I ask for you to show me where the image information differs, you show me a section with the least image information imaginable..

if I take the A77/A57 (The only ones I have to hand right now), and apply the same LR/ACR philosophy (Zero everything for the A57 upsampled to 24MP using bicubic, 13 Chroma Noise for the A77 with black level roughly matched), I get..

Let me zoom in a bit to really see the difference..

And that's 1/500th vs 1/640th for the A77 vs A57 exposures.. I might have slightly overdone the black level matching.. the A77 is a much lighter image due to ISO gains / ACRs default exposure for each camera's RAW.. I can re-do them if necessary..

I don't think you'll ever get the impact of the variability of RAW pre-conditioning (where ever that occurs), so I can see this running and running.. however, I'll happily keep assessing the image information (you know, the thing people actually consume from their photographs), while you blissfully keep posting OOC RAWs (Pentax must love people who do this)..

sensibill
sensibill Veteran Member • Posts: 5,401
Re: A57 or A65 for lowlight wildlife?
2

The A57 is less noisy at high ISO even normalized. Fact. End of discussion.

Nothing else is relevant. IC processing theories, PP theories, pixel level detail at 24MP, which tests are negated according to you for whatever reason. The proof is staring you in the face, and as I said before, I'm done listening to you two grasping for any straw to claim it's otherwise.

Whether OP or others feel the difference is significant should be left to them, but denying it exists altogether with these wildly far reaching wild and nonsensical (dis)qualifiers is a fallacy and disservice to newcomers to SLT.



 sensibill's gear list:sensibill's gear list
Nikon 1 J1 Sony a7 II
philbot Contributing Member • Posts: 846
Re: A57 or A65 for lowlight wildlife?
3

sensibill wrote:

The A57 is less noisy at high ISO even normalized. Fact. End of discussion.

Nothing else is relevant. IC processing theories, PP theories, pixel level detail at 24MP, which tests are negated according to you for whatever reason. The proof is staring you in the face, and as I said before, I'm done listening to you two grasping for any straw to claim it's otherwise.

Whether OP or others feel the difference is significant should be left to them, but denying it exists altogether with these wildly far reaching wild and nonsensical (dis)qualifiers is a fallacy and disservice to newcomers to SLT.



Your image on the right has Chroma NR applied (ACR Default?).. bit of a slip up there..

I'm getting confused as to what you are claiming.. I get the A65/A77 is VISIBLY noisier OOC aspect, I don't understand if you are saying this means the A57 has better high ISO performance as a result (i.e. the image inherently has more information)??

Something doesn't add up, you are hung up on the noise like anyone should care, are you saying we should all get Pentax because it has the least noise by a country mile?


Just to be clear, I'm saying that through the misleading variability of OOC RAW due to pre-conditioning, the A65/A77 actually has identical image information at high ISO..

e.g.

Even with a 28% difference in exposure, the information under the noise from the A77 seems every bit on par with the A57..


I'm still waiting for your rationale for ignoring the difference in RAW pre-conditioning, of which at least one factor is the RAW Converter..

I'll keep quoting DPReview on this,

"Here we look at the RAW files processed through Adobe Camera Raw (in this case version 7.2). Images are brightness matched and processed with all noise reduction options set to zero. Adobe does a degree of noise reduction even when the user-controlled NR is turned off.

The amount of NR applied 'under the hood' is not high, but it does vary by camera (Adobe is attempting to normalize output across different sensors), so inevitably we are still looking at a balance of noise and noise reduction, rather than pure noise levels"


theswede
theswede Veteran Member • Posts: 4,009
Re: A57 or A65 for lowlight wildlife?
3

sensibill wrote:

The A57 is less noisy at high ISO even normalized. Fact. End of discussion.

Except that's in disagreement with the facts. "Noise" is not the same as "chroma noise", or every camera maker would just gaussian blur everything and everyone would be overjoyed at the amazing IQ. Since that's not so, that tells us something.

Namely that the opposite of "noise" is "signal". The measurement of noise is called SNR. Signal to noise ratio. That is how noise is measured. Not by shrill 16MP apologists shouting down anyone with their chroma noise fixated subjective analysis and links to uncontrolled test images.

And here we have the signal to noise ratios:

Thus, at the end of all this, the noise is the same. Unless of course you make the common mistake of thinking "noise" and "chroma noise" are the same thing. But I wouldn't think anyone who has spent some time researching these matters would make such a basic mistake.

Jesper

 theswede's gear list:theswede's gear list
Konica Minolta Maxxum 5D Sony SLT-A37 Sony Alpha NEX-6 Sony a7 Sony a5000 +7 more
sensibill
sensibill Veteran Member • Posts: 5,401
Re: A57 or A65 for lowlight wildlife?
1

It's been said, 'Never underestimate the power of denial.'

Chroma noise is still noise.

The A57 has a chroma noise advantage at higher ISO. This also results in more accurate shadow colors at high ISO.

Chroma noise does not always equate to lost pixel level detail, but NR can reduce pixel detail. This is less evident in brighter or lighter hued areas.

Small variances in exposure do not account for the kind of chroma discrepancies we see in A57 vs A65 at higher ISO. Nor does it render irrelevant what you get from each camera when both are set to ISO-1600 or higher.

The A57 doesn't 'lose' or surrender any shadow detail vs the A65 at higher ISO.

The A65 doesn't have more pixel level detail than A57 when equalized to 16MP.

One A77 DxO test doesn't mean the A65 has equal chroma performance at high ISO.

NR in post processing is lossy. How much so is going to depend on many factors.

Whether the A65 chroma noise will affect your shooting style is subject to individual needs.

Whether the A57 color and chroma performance at high ISO mitigates the other advantages of the A65 (EVF, 24MP, etc.) is subject to individual needs.

People who prioritize chroma performance and shadow hue at high ISO are not stupid, mistaken, wrong, ignorant, rude or dumb. You can both keep voting down my posts, insulting me and doing your level best to shout me down on this, but unlike other former forum regulars, neither I nor the chroma noise are going away.

 sensibill's gear list:sensibill's gear list
Nikon 1 J1 Sony a7 II
theswede
theswede Veteran Member • Posts: 4,009
Re: A57 or A65 for lowlight wildlife?
3

sensibill wrote:

It's been said, 'Never underestimate the power of denial.'

And never have I seen that so strongly as in those die-hard defenders of the 16MP sensor, defying measured reality in their quest to prove their perception trumps any objective measurement.

Chroma noise is still noise.

Naturally. It is one kind of noise. Most sensors will perform rather aggressive chroma noise reduction since many people associate "noise" simply with "chroma noise" and do not comprehend the connection between signal and noise, and how the highly visible chroma noise is only one aspect of how much information a photo contains.

A rather small aspect at that.

The A57 has a chroma noise advantage at higher ISO.

In purely OOC photos this is correct. This advantage in one kind of noise comes at the cost of other types of noise, as evidenced by the SNR ratio being identical.

This also results in more accurate shadow colors at high ISO.

Not correct. The shadow colors are the same. The visible shift is noise, not color. When cleaned up the images look comparable to the point of being indistinguishable, as expected with identical SNR.

Chroma noise does not always equate to lost pixel level detail, but NR can reduce pixel detail.

Which is what happens inside the A57 sensor. Again, as evidenced by the SNR ratio. Signal to noise measurement is not fooled, like our eyes are.

This is less evident in brighter or lighter hued areas.

The relationship is linear. If there is loss of pixel detail it will not be different in different areas.

Small variances in exposure do not account for the kind of chroma discrepancies we see in A57 vs A65 at higher ISO.

And no-one has said this. Chroma noise is one kind of noise. Not the only one, and not even the most important one. Blur noise is much worse at destroying information, which is why NR is so hard.

Nor does it render irrelevant what you get from each camera when both are set to ISO-1600 or higher.

What does this sentence even mean? Of course what we get from the cameras is not irrelevant - it's the whole point of having a camera.

The A57 doesn't 'lose' or surrender any shadow detail vs the A65 at higher ISO.

So you're saying the A57 has higher SNR than the A65. You will have to back that up.

The A65 doesn't have more pixel level detail than A57 when equalized to 16MP.

Whatever "pixel level detail" means. It has the same amount of signal, only distributed differently. That is what matters.

One A77 DxO test doesn't mean the A65 has equal chroma performance at high ISO.

And no-one has said that it does. Can you get through your head that there are more kinds of noise than chroma noise, and that chroma noise is easy to clean up without noticeable loss of information, as opposed to many other more insidious (but less easy to see with the naked eye) kinds of noise?

NR in post processing is lossy. How much so is going to depend on many factors.

Which is why the A65 doesn't do it, and why the A57 has less signal despite having less chroma noise - on sensor NR is still NR, and it is more aggressive on the A57.

Whether the A65 chroma noise will affect your shooting style is subject to individual needs.

The easiest kind of noise to clean up will hardly affect your shooting style much. More insidious sources of noise such as blur noise definitely will. But in the end it's pretty much a toss-up. Click one button or another button to improve the image and you're done, the end result will be indistinguishable.

Whether the A57 color and chroma performance at high ISO mitigates the other advantages of the A65 (EVF, 24MP, etc.) is subject to individual needs.

The A57 and the A65 have images with the exact same amount of noise at high ISO. Your chroma noise fixation is not only tiresome, it is outright daft. There is nothing special about chroma noise warranting this fanatical attention of yours.

People who prioritize chroma performance and shadow hue at high ISO are not stupid, mistaken, wrong, ignorant, rude or dumb.

Yes they are. Ignoring worse problems in favor of easy to spot ones is ignorant at best, dumb at worst. And misleading SLT newbies with shrill shouts of the superiority of a sensor which performs objectively precisely the same is at least rude, and definitely wrong.

There are plenty of reasons to choose a 16MP system over a 24MP system. I know, I did so. But superior high ISO performance is not one of them, and that myth should go the way of the dodo.

You can both keep voting down my posts, insulting me and doing your level best to shout me down on this, but unlike other former forum regulars, neither I nor the chroma noise are going away.

The chroma noise is a simple click to remove. Try replacing detail lost by in sensor NR the same way and we'll see how far you get.

And you make a lousy martyr. You started the name calling, and I don't vote.

Jesper

 theswede's gear list:theswede's gear list
Konica Minolta Maxxum 5D Sony SLT-A37 Sony Alpha NEX-6 Sony a7 Sony a5000 +7 more
philbot Contributing Member • Posts: 846
Re: A57 or A65 for lowlight wildlife?
2

sensibill wrote:

It's been said, 'Never underestimate the power of denial.'

Chroma noise is still noise.

The A57 has a chroma noise advantage at higher ISO. This also results in more accurate shadow colors at high ISO.

I don't know if you are deliberately 'blind' to such blindingly obvious differences between the files..

If you look at the Chroma information in each image, using DPReviews studio comparison tool (to ensure no claims of bias either way), you can see startling differences.

Obviously, I agree, the difference in chroma noise is very much evident, but that's not the only obvious difference.

Look how overall, the A57 colours are much more muted.. the Chroma 'signal' is very much boosted (as is the noise) in the A65..

Now, the first thing that happens when applying Chroma NR is colours become less saturated.. odd that, with the A57 looking less saturated in the comparison tool..


The reason that the A65 cleans up identically to the A57 is simple, the A65 may have more chroma noise, because it has more chroma signal..

Here's ACR's efforts side by side with Chroma NR On/Off.. it actually does a good job at retaining colour fidelity whilst markedly removing the chroma noise.. if you compare it to the A57 image, it has virtually identical detail which ever way you measure it, which can only mean that somewhere along the line (in-camera or in-ACR) the A57 just gets given more Chroma NR.. it certainly has no more signal

zefelder Regular Member • Posts: 198
Re: A57 or A65 for lowlight wildlife?
6

uh-oh.. 16mp vs 24mp again.. looks like it's never going to end!

I guess there are many theories telling why one of these sensors is better than the other for high ISO IQ. There are many ways to compare IQ numerically and numbers will always slightly differ depending on how you compare. And when you look at 200% you can always find some difference in pixels.

But all this absolutely does not matter for real photography. I own both A55 and A77, I shoot a lot with both of them. I do use 3200 and 6400 quote a lot on both, but I yet to find a situation where one camera produces a practically better high ISO photo than the other with the same settings when properly processed (not necessarily with the same settings).

Here are some blind tests of A55 vs A77 @6400 I posted long ago:

http://forums.dpreview.com/forums/thread/3221130#forum-post-41808165

People could not tell much difference between the test shots (even comparing at 100%!), different people preferred different shots, at least not before I told which photo is done by what camera There were several other blind tests posted by others, the results were similar.

sensibill
sensibill Veteran Member • Posts: 5,401
Re: A57 or A65 for lowlight wildlife?
2

Of course, post processed and resized images (well below even 16MP) will show less of a discrepancy. Not a controlled a comparison like DPR, IR or DxO and not really useful for a baseline comparison of OOC RAW.

I keep getting talked down by theswede and philco because DPR has exposure variables in their images, but I've nodoubt they'll heartily agree with your assessment regardless of how you reached it.

For the record, I am not saying one sensor or camera is better than the other, but the fact that the A65 produces more chroma noise at high ISO is not really debatable. Where the discussion becomes convoluted is when we start factoring in theoretical processing and myriad requirements for varied photographers. If all you need is a 1000x1000 web image then sure, some extra chroma noise will matter less, although you will still be left with reddish shadows that need correcting. I guess that 'doesn't matter' either since you can adjust it in post? Regardless of whether that throws overall color balance off and necessitates additional tweaking.

Of course, if all you needed was a 6MP or lower image with smoothing or other processing applied, I have to wonder why we're talking about 16MP or 24MP SLT or SLR level cameras in the first place.

 sensibill's gear list:sensibill's gear list
Nikon 1 J1 Sony a7 II
sensibill
sensibill Veteran Member • Posts: 5,401
Re: A57 or A65 for lowlight wildlife?
2

Chroma noise removal isn't lossless. The discrepancy is not irrelevant for all needs.

Nothing else in your long diatribe is relevant. The personal snark is typical of your style, but also similarly irrelevant. The A65 produces more chroma noise at high ISO.

 sensibill's gear list:sensibill's gear list
Nikon 1 J1 Sony a7 II
Keyboard shortcuts:
FForum MMy threads