A different 'help me choose' thread!

R

Roadkill_6mm

Messages
398
Name
Neil
Edit My Images
Yes
I need help!

I currently own an Oly E-510 with the twin kit lenses and the rather lovely 12-60mm 2.8-4 ED SWD and 50-200mm 2.8-3.5 ED SWD lenses as well as the 35mm macro

I'm looking to switch to Canon.....or Nikon (But I dont want a Canon vs Nikon debate!)

Basically I shoot quite a bit of wildlife - insects and plants in the spring/summer and badgers for most of the year as well as my dogs!. I also now have a 7 month old son so he's been under my lighting for a few photo sessions!

I'm looking to switch away from Olympus for no other reason than low light performance. Unfortunatley where I shoot the badgers is in a forest, so in the summer the canopy is quite thick and it gets dark. Even with the 50-200mm wide open at f3.5 I'm not getting the shutter speeds I need as at iso 800 the noise is very noticable and 1600 is just plain horrible!

The glass which I have would be considered 'semi-pro' and I dont think many will argue with the image quailty they can produce.

What with one thing and another I dont want to plough £1500 into an E-5 and as yet there is nothing in the micro four-thirds which inspires me to continue investing in Oly gear. :shake:

What I need:-

1 x body
2 x lenses!!

Basically I want to replace the 2 main lenses I already have as the focal range gives me everything I want:

12-60mm (24-120 35mm eq) - good all round lens for walking about with and portrait stuff of the lad

50-200mm (100-400 35mm eq) - good for a bit of wildlife, though as ever could probably do with a little more reach

Unfortunatley I know nothing about canon/nikon lenses and what the damn letters mean other than the 'L' glass from canon is good stuff!! And thats before you throw brands like Sigma into the mix - Basically with 4/3rds the 2 SWD lenses I have were the best you were going to get before you're talking proper pro glass at proper pro prices!!

Remember that the Oly has in camera IS and I understand Canon/Nikon this is in the lens?

Finally budget.........

Looking at around £1200 for a body and 2 lenses

I've been looking at the Canon 550d which seems quite good....

So guys, what can you recommend that would fit my budget - remember good performance at high ISO is essential!

Thanks for your help!

Regards

a confused Neil!

P.S. I am happy to take a punt 2nd hand too!!
 
Last edited:
If you want significantly better high ISO performance, you should consider full frame.

The fundamental for low noise is a big sensor that collects a large number of photons in the first place. APS-C is bigger than 4/3rds, but not massively. Full frame is four times the area of 4/3rds, so you're two stops better off there for starters.
 
You'll struggle to get what you want on that budget I think. Full frame either doesn't come cheap or is old and newer crop bodies work better. Bang for buck for wildlife is probably the Nikon d7000 or canon 7d. Both have good AF systems and very decent ISO performance plus cropped so extra reach. Then I'd look at a Sigma 50-500.
 
HoppyUK said:
The fundamental for low noise is a big sensor that collects a large number of photons in the first place.
Click to expand...
:nono: The fundamental for low noise is big pixels on the sensor. Doesn't matter how big the sensor is, it's the size of the pixels that matter (which is why the D3S at 12Mpix and full frame is a low light monster compared to the 21Mpix 5D2)
 
Yeah but the 5dII holds up pretty well against the D3 and D700.
 
Thanks so far guys!

I have considered full frame - but the sheer expense is beyond me,

With regards noise, remember I'm using a 4 year old 4/3rds dslr whose ideal conditions is bright light, not a dark forest!

What sort of lenses should I be looking at am think of putting a bit extra to my budget for a decent zoom for wildlife - budget for that lens alone will be circa £900-£1000 (hopefully!!)

Neil
 
That depends on the system you go for. For wildlife you really need great AF so I'd defo go for one of the models I suggested.
 
There's no point having a 5d when the 7d or d7000 have better high ISO performance and massively better AF systems and give you more reach.
 
micloi said:
And that is why I included the 550D and 120-400mm option :)
Click to expand...

lol - I think that was included as I was replying :)

Must admit I am tempted by the Nikon D7000 - what I might do is get it with a walk around lens now, then wait till the budget is up a bit more to get some decent glass in the 300mm zoom range

What do I look for in Nikon lenses?? I know for Canon 'L' glass would be what I'm after.

And am I right in thinking the Nikon's crop factor is 1.5?

Neil
 
Yes, 1.5. With Nikon you're really looking at the fixed aperture lenses as a mark of quality, but there are some gems like the 70-300 VR too.
 
The23rdman said:
Yes, 1.5. With Nikon you're really looking at the fixed aperture lenses as a mark of quality, but there are some gems like the 70-300 VR too.
Click to expand...

I think the fixed aperture lenses are going to be waay out of my price range unless I win the lottery!!

Just to add to the mix is there much difference between Canon and Nikon lenses in the sub £1000 price range in terms of focusing, sharpness and speed?

Neil
 
I don't know the Canon offerings but the Nikon 16-85VR is the equivalent to your 12-60, though a bit slower. It is my most used, general purpose walk-around lens and I am very pleased with it. I also find the 70-300VR and excellent longer zoom.

Both these lenses focus fast and accurately on the Nikon bodies. Add them to a D90 and I think you would be within your budget if you got some of them second hand.
 
arad85 said:
:nono: The fundamental for low noise is big pixels on the sensor. Doesn't matter how big the sensor is, it's the size of the pixels that matter (which is why the D3S at 12Mpix and full frame is a low light monster compared to the 21Mpix 5D2)
Click to expand...

No, sensor area is the starting point, not pixel size. Canon 40D and 5D2 have similar pixel size, but the 5D2 has twice as many of them and much better noise performance as a result.

It matters far less how you slice and dice the sensor with pixels. Sure, if you have high pixel density then some of the usable sensor area is taken up with the gaps between them, but that's overtaken by sheer area. 5D2 is no slouch on the noise front and blitzes any APS-C camera, plus it has very high resolution compared to the Nikon D3S.
 
HoppyUK said:
No, sensor area is the starting point, not pixel size.
Click to expand...
No, a small pixel will always be inherently more noisy than a larger pixel as the larger pixel will be able to capture more signal (light) to the noise generated in the chip assuming all else stays the same (sensitivity of the substrate to light and noise induced in it) for a given shutter speed and aperture. That's just physics....

What you are saying is that given the same framing, the larger format and effectively higher pixel density (if the pixels are the same size, the pixel density is the same but as you have more covering the same area, the effective - i.e. area of the image covered by each pixel - density is higher) means you can hide the noise better as each noisy pixel contributes to a smaller area of the resultant image.
 
Andy, with respect, you are contradicting yourself in your last line - to 'hide the noise better' is kind of saying the same thing. Noise is the system's contribution to the image, compared to the signal. It doesn't matter if you collect that signal in one large bucket or several smaller ones. Provided the same total signal is collected, you will end up with the same signal-to-noise ratio.

The comparison with a 40D and a 5D2 is a good one in this respect as they have very similar pixel density and from roughly the same generation of sensor technology. And the 5D2 has far less noise, because it has twice the number of pixels. In Nikon terms, compare the full frame D3 (12mp) with the D3X (24mp) and while the D3 edges ahead in the noise stakes, the D3X is not very far behind, and I would suggest that the difference it because the D3X's smaller pixels waste a bit more light that falls between the 'buckets'.

Sure there are other differences and every new generation takes things a step forward, but to simply say that pixel size defines noise clearly doesn't stand up in practise, whereas total sensor area pretty much does.
 
HoppyUK said:
Andy, with respect, you are contradicting yourself in your last line - to 'hide the noise better' is kind of saying the same thing. Noise is the system's contribution to the image, compared to the signal. It doesn't matter if you collect that signal in one large bucket or several smaller ones. Provided the same total signal is collected, you will end up with the same signal-to-noise ratio.
Click to expand...
Sorry, that's wrong. You will get more signal and less noise in a larger pixel when compared to a small pixel for a given technology (nothing to do with edge effects as microlensing has eliminated that now). That's why larger pixel cameras are better at taking higher ISO images in the same technology and why compacts are rubbish at high ISO. A 10 MPix sensor in compact, crop and full frame (assuming the same technology) will produce high ISO performance in the order compact < crop < full frame. Yes, the bigger sensor is better, but that's due to the pixel size, not the sensor size.

HoppyUK said:
Sure there are other differences and every new generation takes things a step forward, but to simply say that pixel size defines noise clearly doesn't stand up in practise, whereas total sensor area pretty much does.
Click to expand...
Really? Which is going to have better noise performance - a 20Mpix crop or 30Mpix full frame in the same technology? Without working out the pixel size of each, you have no idea.....
 
arad85 said:
Sorry, that's wrong. You will get more signal and less noise in a larger pixel when compared to a small pixel for a given technology (nothing to do with edge effects as microlensing has eliminated that now). That's why larger pixel cameras are better at taking higher ISO images in the same technology and why compacts are rubbish at high ISO. A 10 MPix sensor in compact, crop and full frame (assuming the same technology) will produce high ISO performance in the order compact < crop < full frame. Yes, the bigger sensor is better, but that's due to the pixel size, not the sensor size.

Really? Which is going to have better noise performance - a 20Mpix crop or 30Mpix full frame in the same technology? Without working out the pixel size of each, you have no idea.....
Click to expand...

That's easy - the full framer (if it existed).

I have given examples that prove pixel size is not the defining factor, and that sensor area is. Can you give one example of a crop format camera that has better noise performance than an equivalent technology full frame? Even disregardling pixel size, there isn't one - not by some margin.

And gapless mircrolenses (not that they are truly gapless) is not the same as gapless pixels.
 
Richard,

You're flying against everything that is known from physics. Look at: http://en.wikipedia.org/wiki/Image_noise

I'll quote from that article:

Wikipedia said:
or effective light collection area per pixel sensor, is the largest determinant of signal levels that determine signal-to-noise ratio and hence apparent noise levels
Click to expand...

What you are missing is that most noise apparent in the sensor isn't related to area - it's related to quantum physics and temperature.......
 
Blimey Andy, just read the link you have posted.

In the section headed "Effects of sensor size" it says: "...the pixel count makes little difference to perceptible noise levels &#8211; the noise depends primarily on sensor area, not how this area is divided into pixels."

Which is exactly what I said above. Without the need for any quantum physics, it's pretty obvious that the more light you've got to play with (however it's collected) the less it needs to be amplified, so you get less noise.
 
HoppyUK said:
Blimey Andy, just read the link you have posted.

In the section headed "Effects of sensor size" it says: "...the pixel count makes little difference to perceptible noise levels – the noise depends primarily on sensor area, not how this area is divided into pixels."

Which is exactly what I said above. Without the need for any quantum physics, it's pretty obvious that the more light you've got to play with (however it's collected) the less it needs to be amplified, so you get less noise.
Click to expand...
Richard,

The sensor area refers to the individual pixel size. Why else do you think that a D3S and 5D2 have differing noise values (D3S being better). One has 12Mpix, the other 21Mpix. Not surprisingly, the differentiator is the size of the pixels, not the size of the sensor....
 
PS. I really can't see how this could be clearer...

Wikipedia said:
For images at lower signal levels (higher ISO settings), where read noise (noise floor) is significant, more pixels within a given sensor area will always make the image noisier.
Click to expand...

I.e. pixel size is the differentiator...
 
arad85 said:
Richard,

The sensor area refers to the individual pixel size. Why else do you think that a D3S and 5D2 have differing noise values (D3S being better). One has 12Mpix, the other 21Mpix. Not surprisingly, the differentiator is the size of the pixels, not the size of the sensor....
Click to expand...

You are mistaken. When it says sensor (in your link) it means sensor, not pixel. To repeat the quote, "...noise depends primarily on sensor area, not how this area is divided into pixels." If you want to split hairs, refer to the word 'primarily' which in the context of full frame vs APS-C is 'overwhelmingly'.

I've explained the main reason for the D3S's superiority, which compared to the D3X (which is a more direct comparison) is slight. Check the DPReview tests for the D3X where they have images and graphs side by side. Also look at the D300S where it is set against a Canon 7D (with 50% more pixels, and crammed into a slightly smaller sensor) where it says "the higher resolution 7D would also trump the Nikon in terms of overall image noise." How does it manage that if pixel size is the main factor? There are any number of examples like this, but none which support your contention.

arad85 said:
PS. I really can't see how this could be clearer...
Click to expand...

Quite. As explained in your link.

[I.e. pixel size is the differentiator...
Click to expand...

No. See above. More light collected means less amplification which means less noise. It's simple and obvious common sense.
 
Is there an award for the most complete hijack of a thread? It's amusing reading your debate about sensor noise, but you're not really helping the OP choose a camera system.

Back on topic, I'd suggest that if you're serious about wildlife then the Canon range offers you lots more lens options than Nikon. Both manufacturers have a very decent 70-300 with IS/VR for around £400, but if you want anything better then the Canon triumvirate of 100-400mm L IS / 300mm f/4 L IS / 400mm f/5.6 L blows away the Nikon equivalents.
 
Stewart

Thanks for getting the thread back on topic!!

Thus far I've heard mainly about Nikon - body wise was looking at something like the D90, or hopefully at a push the D7000.

Factoring I the cost of a 70-300mm lens, I'll be left circa £700 ish for a body. Was thinking regarding a walkabout lens I hear good things about the 'nifty-fifty' which may well suit my needs for the time being, even though I prefer the flexibility of a zoom.

So with regards Canon, what sort of body should I look at in the £700 range??

Neil
 
Richard - on rereading the links (and in absence of the alcohol I was drinking whilst watching the rugby ;)) the bit we missed out was:

Wikipedia said:
In the case of images bright enough to be in the shot noise limited regime, when the image is scaled to the same size on screen, or printed at the same size, the pixel count makes little difference to perceptible noise levels &#8211; the noise depends primarily on sensor area, not how this area is divided into pixels.
Click to expand...
I read that as when properly exposed, a larger sensor will produce less noise when the image from it is scaled to a given size. This makes sense as you are dominated by signal rather than sensor noise (shot noise is variations in the light hitting adjacent pixels so is to do with the incoming signal, not the characteristics of the sensor) and with the larger sensor, you are scaling the image less.

It then follows with:
Wikipedia said:
For images at lower signal levels (higher ISO settings), where read noise (noise floor) is significant, more pixels within a given sensor area will always make the image noisier.
Click to expand...
I read this as, when you are fighting for light and pushing the ISO to get the exposure, pixel area is the dominant factor. That's exactly what I've been saying.

As to the OPs questions, for high ISO capabilities of cameras, have a look around dpreview as they hold comparative tests for different cameras. Have a look at the raw and JPEG comparisons here for example of the 550D vs others: http://www.dpreview.com/reviews/canoneos550d/page14.asp

Some other points to think about:
- if you limit yourself to ISOs below about 800, all cameras will produce much the same image as the ISO performance is near enough identical.
- it may be better to get a camera with slightly worse overall noise performance but that has higher ISO available if you really are pushing things often. Sometimes getting the shot is more important than how noisy it is.
- If you're shooting JPEG, the way the camera processes the raw data will influence the look of the picture more than the inherent noise in the chip. Take a look at the comparisons on dpreview for example. In particular, look at the differences between raw noise and the JPEG saved at that ISO.
- Lens choice should be a consideration in the body you choose. For low light, you really want low f numbers. What lenses are available at a suitable cost? (for lowish cost on the Canon, I'd be looking at the Tamron 17-50 f2.8 NON VC for example). If you're looking at lenses on Canon bodies, there's a great set of comparisons at The Digital Picture see: http://www.the-digital-picture.com/reviews/iso-12233-sample-crops.aspx Canon generally have a bigger range of lenses that tend to be slightly cheaper than Nikon equivalents for example.
- Be aware that higher Mpix counts will "demand" better glass as they have more resolving power than lower mpix count cameras. Whether this demand worries you or not is a personal thing
- To find out lens/body prices, take a look at http://camerapricebuster.co.uk
- If you do buy one of the physically smaller bodies, be careful that putting the larger and better lenses on it can make it feel like a lens with a camera attached rather than the other way around.

If you do go Canon, the 550D, 600D, 60D and 7D will all produce near enough (exactly??) the same image. They all share a the same 18MPix sensor and you are paying for different features in the lineup. It's not clear what the image quality of the 1100D (which is a 12Mpix sensor) will be like as it's not been released.

At £700, I'd be looking at the 60D in the canon range as it's the first to have a decent body size (IMHO).
 
Last edited:
Andy - cheers for that, given me food for thought.

With regards limiting myself to iso800 I actually limit myself to iso400 where possible as I'm not happy with the noise created in the circumstances by my E510 at that level and iso1600 (which is it's limit) is a damn sight worse.

My thoughts about going canon/Nikon is that the sensor is bigger (above argument taken on board!!) and a more modern camera should be able to handle it a bit better!!

Regards

Neil
 
Just a quick answer to an earlier question, there is no nikon equivalent designation to canon's 'L' glass, just have to do your research annoyingly! The 70-300 and 16-85 are probably the top "consumer" telezoom and standard zoom for crop sensor cameras (tho the 70-300 is a full frame lens).

I have both of these lovely lenses and a D90, and am so far not tempted by the D7000: doesn't seem like twice the camera, but is nearly twice the price of the D90. A recent test of the D90 vs D7000 in one of the monthlies showed marginal differences in noise between the two, and as that seems to be one of your prime factors, I'd save your cash.

Should you go down the nikon route, obviously ;)
 
Last edited:
Roadkill_6mm said:
Andy - cheers for that, given me food for thought.

With regards limiting myself to iso800 I actually limit myself to iso400 where possible as I'm not happy with the noise created in the circumstances by my E510 at that level and iso1600 (which is it's limit) is a damn sight worse.

My thoughts about going canon/Nikon is that the sensor is bigger (above argument taken on board!!) and a more modern camera should be able to handle it a bit better!!

Regards

Neil
Click to expand...

Sorry about the noise stuff, but it is relevant to your OP.

Looking at the options, you can do a lot with noise reduction software these days. Neat Image and Noise Ninja for example. There are one or two posters on here that get amazingly good results with a bit of careful post processing.
 
You must log in or register to reply here.
Back
Top