Canon 5D Mark iii

[koffiejunkie]

One more thing, since you mention macro photography. You should be aware that the same lens at the same aperture will give you less depth of field on a full frame body. Most people see this as a positive but in macro it definitely isn't.

Anyway, as others have said, we need to know which lenses you have.

 

[fvdbergh]

While I cannot comment on the 1000D I do have both the 400D, which is almost identical to the 1000D when it comes to ISO, and the 7D and the difference between the two is staggering. First off the 1000D/400D has a max ISO of 1600 where as the 7D goes to 12800. In real terms there is at least a two stop improvement, if not more, when it comes to using that higher ISO.

I have not looked at the image quality difference in great detail, so I'll defer to those with actual experience with both the 1000D and the 60D, for example. When I said there is a 0.5 stop difference between the 1000D and the 60D, I was referring to the dynamic range figures posted on sensorgen.info --- not the most reliable source, I'll admit. Still, it is unlikely that the real world dynamic range difference would be 1 stop; it is likely to be closer to 0.5 stop.

Just be careful to not confuse the manufacturer's max ISO with actual performance. On a digital sensor, the ISO setting is just the gain on the amplifier between the photodetector (sensor element) and the analogue-to-digital converter. The noise (photon shot noise, specifically) is amplified along with the signal, with the relationship becoming very close to linear at higher ISO settings. For example, take a look at Bill Claff's charts. Select the 7D. Notice how the dynamic range at ISO 3200 is 5.04 bits (stops), followed by ISO 6400 with 4.04 stops, and ISO 12800 at 3.08 stops. For every doubling in ISO (hence doubling in amplification of the analogue signal), the actual dynamic range (difference between noise level and signal level) halves (i.e., decreases by 1 stop).

Now for the tricky part: notice that the actual dynamic range is quite small compared to the ADC resolution, e.g., at ISO 3200 we only use about 5 bits out of the available 14 bits (or is it 12 bits on 7D?). This tells us that we can simulate higher ISO values by underexposing on purpose, and pulling up the brightness in post. Try the following experiment: Set the camera to ISO 6400, and meter a scene in aperture priority. Take a shot, switch to ISO 3200 and manual mode. Use the same shutter speed that the camera metered for the ISO 6400 shot, take the shot. Now open the raw image of both the ISO 6400 shot and the ISO 3200 shot in your favourite raw editor. Pull up the underexposed ISO 3200 shot by one stop, which should give you an image that has the same apparent brightness (exposure) as the ISO 6400 shot. You should see that the apparent noise is almost identical (although at some point the quantization noise will start to dominate).

So why do manufacturers enable higher ISO settings on newer bodies? Take the 700D, which now has an 12800 setting that the 600D did not have. We know that the sensors on the 700D and the 600D are virtually identical (except for the on-sensor phase-detection AF sensors introduced with the 650D). But the 700D has the new Digic sensor (V versus IV, I think?). So what has happened is that they use the increased processing power of the new processor to implement better (but slower) noise reduction algorithms on the 700D. This allows you to pull more usable ISO 12800 JPEG shots from the 700D, compared to the 600D which had less advanced noise reduction algorithms (speculation on my part, I must admit), hence they limited the 600D to ISO 6400 (or whatever). But this only applies to OOC JPEG shots

My experiment above should be repeatable across the 600D and the 700D, i.e., if you shoot in raw, then a 1-stop underexposed ISO 6400 shot on the 600D, pulled up by 1 stop in post, would be nearly identical to an ISO 12800 shot on the 700D.

The situation becomes a little more complicated on some brands, e.g., Pentax does some noise reduction even on raw files, but as far as I know neither Canon nor Nikon tamper with the raw files (other than hot-pixel suppression).

The 60D/7D also uses the newer Digic IV processor (dual processors in the 7D) as compared to the 1000D's Digic III.

This does not affect the quality of raw files, only JPEGs. See argument above.

Recommending fast glass is great but shooting wide open because you have to is not the same as doing it because you want to. A razor thin DOF should be an artistic choice, not an necessity. That's where the improved high ISO of the newer bodies comes into play

No argument there. I guess my main point was that you can explore a fast prime cheaply (~R1k), and see if that suits you.

 

[fvdbergh]

I'll echo this. I had one for two years and I can highly recommend it.


I'll also agree with what fvdbergh said about noise performance. The difference is slight. I don't find the level of noise present to be much different between the 7D and 5DmkIII (there is at least 3 years between those two cameras). BUT one thing that is different is the colour retention/reproduction. That's very different.

Similarly, the 7D wipes the floor with everything that came before it when it comes to the colour it retains when shooting at, say 3200 ISO. Even the 50D doesn't measure up to it (I had that one too).

Don't confuse noise with image quality. Noise can be dealt with, lack of image cannot.

Well, it is very hard to characterise image quality with a single figure; the same goes for noise. The hardest part is that no single one of the objectively measured statistics agree 100% with subjective human perception of noise.

This is especially true when comparing the noise in the Red, Green and Blue channels. For example, some people find colour noise more objectionable than others. Mix in some trickery (google 500D colour blindness) that the manufacturers use over time, and you will be able to obtain quite different subjectively perceived image quality between two bodies that use (mostly) the same sensor.

My guess would be that the 7D (and 5DIII) use better colour filters, i.e., closer to the ideal spectral response. This could explain why 7D images might look better than, say, 500D images, even though they use nearly the same (or is it exactly the same) CMOS sensor behind different colour filters.

 

[koffiejunkie]

My guess would be that the 7D (and 5DIII) use better colour filters, i.e., closer to the ideal spectral response. This could explain why 7D images might look better than, say, 500D images, even though they use nearly the same (or is it exactly the same) CMOS sensor behind different colour filters.

The 500D and 50D uses an older 15MP sensor, the 7D, 60D and 550D onwards all have the newer 18MP sensor, that Canon seems to have decided to sick with.

My comment was based particularly on how much accurate colour I see in high-ISO shots, particularly what objects in the background (and often underexposed) look like. I've owned the 40D, 50D, 7D and now 5DmkIII. The difference beween the latter two is substantial but much less than I expected. Between the 40D and 50D, the difference was substantial - I never used 3200 on the 40D while on the 50D I was quite happy to.

When I was considering the 7D, this was the biggest consideration. I was just about to go on holiday and spend some time taking night-sky and night-scene pictures (star-lit, moon-lit kinda stuff, so pushing the limits slightly). Someone I know who owned both the 50D and 7D at the time took some shots for me to compare. Same settings on both cameras, same scene and composition, same time, and sent me the CR2 files. This was specifically 3200 and 6400 ISO, as I was looking at using shorter shutter times to avoid star trails. One was of a farm yard, barn in the background, fence, sleeping dog, trailer with hay on, that sort of thing. The difference in colour between the two was staggering. The 50D's image looked like a bad tone-mapping HDR job by comparison.

Below 3200 there's little to choose between the camera, but the way pictures look, and the way colour is reproduced at high ISO definitely sets them apart. I have not yet seen a review or a measuring system that accurately reflects this.

 

[bwana]

Right now I'm looking at a series of images taken with the 400D and the 7D and as far as my eyes are concerned the noise levels of the 400D at 1600 are the same as the 7D at 4000/5000 ISO. I really don't need much more than that to say there is a significant difference between the ISO capabilities of the two generations of cameras. IOW The newer camera yields far better results.

 

[fvdbergh]

The 500D and 50D uses an older 15MP sensor, the 7D, 60D and 550D onwards all have the newer 18MP sensor, that Canon seems to have decided to sick with.

I should have spent the 10 seconds doing a quick google. Ah well.

Below 3200 there's little to choose between the camera, but the way pictures look, and the way colour is reproduced at high ISO definitely sets them apart. I have not yet seen a review or a measuring system that accurately reflects this.

DxOMark does a "Colour response" test, which should be able to give you some additional information. Unfortunately, I do not currently know enough about colour perception to interpret those results completely. Comparing the 50D and 7D does reveal that the 7D has a more selective red channel than the 50D (which really measures green and red equally weighted in the red channel), which may explain why night-time colour (typically low in red light?) would be more accurately captured by the 7D.

... but this is specifically relative to the sRGB primaries, which are probably different from a human visual response.

It may also be that more expensive bodies use more efficient Bayer colour filters, so that more light reaches the sensor --- but now I am just speculating

 

[fvdbergh]

Right now I'm looking at a series of images taken with the 400D and the 7D and as far as my eyes are concerned the noise levels of the 400D at 1600 are the same as the 7D at 4000/5000 ISO. I really don't need much more than that to say there is a significant difference between the ISO capabilities of the two generations of cameras. IOW The newer camera yields far better results.

Yeah, that is my mistake. I incorrectly assumed that the 1000D also used the 18 MP sensor that the rest (550D upwards) is using, so there may be a real difference between the 15 MP and the 18 MP Canon sensors. My comments thus only apply to all the 18 MP bodies, i.e., between them, there are no great improvements in dynamic range.

 

[koffiejunkie]

It may also be that more expensive bodies use more efficient Bayer colour filters, so that more light reaches the sensor --- but now I am just speculating

Or maybe more efficient/accurate debayering algorithms (or is that the same)?

Whatever they're doing, I like the results

Now, if Canon can only get rid of the red bands...

 

[bwana]

Or maybe more efficient/accurate debayering algorithms (or is that the same)?

Whatever they're doing, I like the results

Now, if Canon can only get rid of the red bands...

How is everyone going to know you paid a lot for your lens?

 

[koffiejunkie]

How is everyone going to know you paid a lot for your lens?

Ha ha! Those are not the bands I was referring to

Speaking of lens, since getting the 5D, I'm loving my 24-105L and 50mm f/1.4 more than ever before.

 

[bwana]

Ha ha! Those are not the bands I was referring to

Speaking of lens, since getting the 5D, I'm loving my 24-105L and 50mm f/1.4 more than ever before.

Still not really interested in the 24-105 myself and tbh I rarely use my 50mm anymore

 

[koffiejunkie]

Still not really interested in the 24-105 myself and tbh I rarely use my 50mm anymore

Not exactly the ideal pair of lenses to shoot cricket with!

 

[Polymathic]

right i can only dream of getting a L lense

 

[OGroteKoning]

Okay here are the lenses :
Sigma 70-300mm 1:4-5.6 APO DG
Tamron SP 10-24mm 1:3.5-4.5 Di II
Sigma 70mm 1:2.8 DG Macro EX
Canon EFS 18-55mm (came with camera)

... and the flash
Canon Speedlite 430EXII

 

[OGroteKoning]

If the visual difference between EF and EFS lenses are the bit to the right of the red line, then only the lense that came with the camera is an EFS lense (Canon EFS 18-55mm)

But please comment

 

[fvdbergh]

Okay here are the lenses :
Sigma 60-300mm 1:4-5.6 APO DG
Tamron SP 10-24mm 1:3.5-4.5 Di II
Sigma 70mm 1:2.8 DG Macro EX
Canon EFS 18-55mm (came with camera)

... and the flash
Canon Speedlite 430EXII

Ok, so I see you have the flash --- I would recommend looking into remote triggering solutions as an interesting addition to your wife's macro photography (others should chip in regarding pocketwizards, or whatever solution they think is best).

But I also see you do not have any f/1.8 lenses. You will get a boost in low light performance at shorter focal lengths with a 35 mm f/1.8 prime lens, and you can throw in the 50 mm f/1.8 prime lens for an additional R1k. As bwana pointed out, there are disadvantages to using a faster lens (lower f-number), most important of which is that your depth of field is shallower, which may be desirable in portraits (for some), but not necessarily for all types of photography. But for the price of a 35 mm and a 50 mm pair of prime lenses, this would be an interesting option to consider.

You could also consider a constant aperture "short zoom", like the Sigma 17-50 mm f/2.8 or its Tamron equivalent, which will set you back around R8k (less for the Tamron, I think). This lens will give you an f/2.8 aperture over the roughly the same range as your kit lens. The main problem with this option is that it is not that much faster at the 17 mm end (kit lenses are usually f/3.5 there), although it is much faster at the 50 mm end (f/2.8 vs f/5.6 on the kit lens). If you upgrade the body (say to a 650D) this option becomes more attractive ... but still much more expensive than just buying the two prime lenses I mentioned above.

Anyhow, could you give us some hints of what types of photography your wife prefers (other than macro)? Portraits, landscapes, action (sports, kids, whatever)? Specifically, for which type is the low-light performance currently the problem?

 

[bwana]

Okay here are the lenses :
Sigma 60-300mm 1:4-5.6 APO DG
Tamron SP 10-24mm 1:3.5-4.5 Di II
Sigma 70mm 1:2.8 DG Macro EX
Canon EFS 18-55mm (came with camera)

... and the flash
Canon Speedlite 430EXII

Sigma 60-300mm 1:4-5.6 APO DG - presume you mean 70-300? Not sure if that one is EF or EF-s
Tamron SP 10-24mm 1:3.5-4.5 Di II - EF-s - the mount might physically fit an EF body but there will be severe vignetting if so.
Sigma 70mm 1:2.8 DG Macro EX - EF lens so you can use that with both types.

 

[OGroteKoning]

@fvdbergh

She likes general photography - mostly in an art-like way.

Macro photography
Stills (very arty!!!)
Nature photos at any given level of light
Wildlife photos
People (yup - last on the list)

She is not that big into action photos, but obviously with the equipment we have, we should be able to cope with any action photo we want to take

It's not like she has specific problems with specific low light photograph - it is more general. She would love to take photo's of our baby without using the flash.

She has coped with the a cheap tripod for longer exposures, but what a schleb to set up. In darker environments, she used the flash often enough, but to me that really "flattens" a photo - to me it takes away natural depth (hope you understand what I am trying to say

 

[OGroteKoning]

Sigma 60-300mm 1:4-5.6 APO DG - presume you mean 70-300? Not sure if that one is EF or EF-s
Tamron SP 10-24mm 1:3.5-4.5 Di II - EF-s - the mount might physically fit an EF body but there will be severe vignetting if so.
Sigma 70mm 1:2.8 DG Macro EX - EF lens so you can use that with both types.

Sorry, yes the Signma is 70-300 (fixed now)
How do you know if the lenses are EF or EFS? Where do you look it up?

 

[bwana]

Sorry, yes the Signma is 70-300 (fixed now)
How do you know if the lenses are EF or EFS? Where do you look it up?

You can tell either via reviews or from looking at the manufacturers websites.

She is not that big into action photos, but obviously with the equipment we have, we should be able to cope with any action photo we want to take

Action photography, especially sports, usually requires more specialised equipment - especially in the lens department.