Help choosing a dSLR-like (megazoom) - disappointed by Canon's SX series!

[longtalker]

Hello everyone,

I want to buy a camera that has the following features, which my 6 year old Canon A610 does not have:
- very large optical zoom (at least 20x)
- high res video (at least 720p, ideally 1080p)
- exposure bracketing
- articulated LCD (the A610 does have that one)

After a long comparison of bridge ("dSLR-like" compacts) and dSLR cameras, I eliminated the latter category because adding a long zoom lens to the price of a dSLR is something I cannot afford. After comparing the available bridge cameras, I decided that either the Canon SX1 or the SX30 meet my requirements the best.

However, before proceeding to decide between these two (essentially, a choice between the SX1's 1080p filming and higher fps in burst mode and the SX3's better optical zoom and higher pixel count), I could not get over how disappointing the image quality in both of these two cameras is. The samples that I've seen from both of them seem to almost always have a soft/watercolour aspect to them, regardless of the focal length used, and even at the lowest ISO and in bright sunlight. These samples are noisier, and the details at 100% crop are much less clear/crisp than those produced by my old Canon A610! It seems just silly to think that I would take pictures with my old camera when I want quality images and that I'd use the new one when I need lots of optical zoom and/or HD videos.

Could this difference in IQ be attributed just to sensor size? Surprisingly, the A610 has a larger sensor (1/1.8") than the SX1 or SX30 (1/2.3"), even though it has a much smaller resolution (5 MP).

I also expected there to be a large difference in IQ between the SX1 and the SX30 because the former has a CMOS sensor while the latter has a CCD one (I expected the CCD to be better, but either way, I expected *one* to be better than the *other*) - but the images are just as disappointing for both of them.

I'd be willing to sacrifice the articulated LCD for a mega-zoom camera that has an image quality the same as my old A610, but after reading reviews and comparing samples, it seems the SX1/SX30 are really as good as it gets in their category.

My question is, therefore: I know that megazooms don't do well with poor lighting, but are they really unable to produce a good IQ even with enough light? Should I just look for sensor size? Would I have to get a dSLR if I wanted good IQ (if "good" is the standard set by a 6 year old point&shoot!!) and only gain the exposure bracketing and HD filming from my wish list of new features? Could people please recommend what cameras I should consider given all my criteria above?

Many thanks in advance for any help!

 

[Dave in Wales]

Welcome Catalin,

Sounds to me as if you are searching for the 'Holly Grail' of cameras.

If all you asked for and compatable IQ to a DSLR then we'd all be using them, I sure would, why carry a bag of gear when one camera will do it all.

It's all a trade-off, a sacrifice, what you gain on the swings you loose on the roundabouts.

'Jack of all cameras(trades) master of none'

Take a look at the Canon G12, it falls short on your requirements, but IMVHO is the best compact available today.
I have an old G7, the IQ is outstanding.

D in W

 

[dejongj]

Also what is 20x optical zoom? What actual range are you trying to cover? The combination of sensor size and what you are trying to achieve would become of importance as well.

And yes you generally sacrifice IQ by zoom, you can negate it a little by spending lots and lots of money.

But the ultimate question is, what range are you looking for to get? A 20x optical zoom doesn't mean anything unless you know what you are starting from.

For example a Nikon D3100 entry level dSLR with a Tamron 18-270 PZD (£500 for just the lens) only give you a 15x zoom....

 

[srichards]

New sony bridge is supposed to be pretty flipping good and also panasonic have brought out one with a longer zoom than their previous offerings:

http://www.photographyblog.com/reviews/sony_cybershot_dsc_hx100v_review/

http://www.photographyblog.com/reviews/panasonic_lumix_dmc_fz150_review/

Both of them are about £400 so much cheaper than a dslr with the needed zoom range.

 

[longtalker]

Thank you for your replies! I completely agree about it not being realistic to expect a camera that has all the "pros" of the various categories of cameras without any of the "cons" that come with them. However, I really just wanted to have the IQ of a much older&cheaper camera (the A610), and was wondering whether SX1's and SX30's failure to match even that standard is due to their small sensor sizes, or to their long focal range lenses.

Sorry to not mention the range of the lenses. Both the 20x lens of the SX1 and the 35x lens of the SX30 are meant to take you from quite a "wide" focal length - 28 mm (equiv.) for SX1 or 24 mm for SX30 - all the way to the telephoto suggested by their respective optical zoom value. I would be happy to have the SX30's extra 4 mm for "wide", but it might come at a cost of significant geometric distortions; from the samples, however, I see that both cameras show visible barrel distortion at the wide end, with no real difference between them.

 

[squishy]

A massive zoom range and a small sensor is simply asking too much of a camera IMO.
You're demanding a lens with such extreme parameters to resolve detail onto a tiny space

One thing you shouldn't do though is compare 100% crops with your old 5mp A610. The SX1 and SX30 have twice and nearly 3 times the MP number, so you're taking a lot smaller area of image to look at. If you want a fair comparison, downsize those images to 5mp and see how they look.

Those new sony and panasonic super-zooms do look very impressive given the specs and image quality.

 

[HoppyUK]

I don't know if this review will be helpful, but it's pretty comprehensive and covers six cameras around what you're looking for http://www.dpreview.com/reviews/q311travelzoomgrouptest/

I'd be surprised if modern cameras were not at least as good as your older one, even if they're not dramatically better. The small size of the sensor is the main thing that limits image quality, but it allows you to have a very wide range zoom in a small package.

To get the same sort of range with a lens for a larger format camera with better image quality, you're immediately into much greater size and cost.

It does sounds like you are seeking the Holy Grail. You should get to try one of these hyper zooms first - even if you get what you want on paper, it may not deliver in the way you want/expect. They're not great to use at the long end.

If that turns out to be the case, decide what it is that you mainly want to do, and address that first, then build around it as best you can. Buit it's no surprise that photographers who like to cover a wide range of stuff in all conditions tend to have a lot of gear

 

[GardenersHelper]

I have slightly strange requirements because most of what I do is close-ups/macros, so these thoughts might not be particularly relevant, but FWIW I have been pondering similar issues for a couple of years now.

I used a Canon S3is and then until very recently an SX10is. I was looking for an upgrade to image quality and the means of working better in often poor light levels as I much prefer to use available light rather than flash.

As for you an articulated screen was important for any upgrade, and I wanted zoom covering at least around the range provided by the SX10. I don't use exposure bracketing much, and never use video and don't know anything about different cameras' video capabilities.

Better dynamic range (with less tendency to blow highlights) was highly desirable for me. I am used to dealing in PP with the noise of the S-series cameras, especially above ISO 200 or so, but I really wanted a bit of an improvement on that front.

By looking at comparison shots at a couple of sites I couldn't convince myself that an SX20 or SX30 would really be an upgrade in terms of quality - a very slight downgrade if anything. None of the other superzooms ticked all my boxes. The Fujifilm HS20 came close and seemed quite promising on the IQ front, including its clever tricks for extending dynamic range, but the screen articulation made it unsuitable for me.

I pretty much gave up on improving things by using a superzoom and eventually upgraded to a Panasonic G3 last month, but of course this is a more expensive option.

Panasonic have since announced the FZ150 (with a reduction in pixels from the IQ-challenged but otherwise frustratingly nice FZ100), and the initial sample photos look quite promising, promising enough in fact that when more feedback is available I will consider getting an FZ150 to cover some requirements that my G3 kit doesn't handle well rather than spending even more on additional G3 kit that might not solve the problems anyway. Had the FZ150 been announced just a little earlier, I might well have gone for that rather than the G3 (but my requirements are untypical).

All that said, if you are inclined to look at 100% crops of images then I doubt any of the current crop of superzooms with 1/2.3" sensors is going to please you, especially above base ISO (and to get a 20x or more zoom range while keeping the camera to an acceptable size/weight you do apparently need a pretty small sensor and all that implies for IQ and the use of higher ISOs).

If though you look at whole, or not too radically cropped images, and view on screen at perhaps 1200 to 1600 pixels across, or print at perhaps 12" or so along the long side, then the FZ150 might be worth looking at (incidentally, unlike the Canon superzooms, you can use RAW with the FZ150, although given the very small sensor this might not actually be much of an advantage). I have found it quite informative to lurk in the dpreview Panasonic Talk forum, and also their Canon Talk, Fujifilm Talk and Sony Talk forums, which is where superzooms are discussed, analysed, argued over, ranted about etc.

 

[longtalker]

Hi Adam, Richard and Nick, and many thanks for your very helpful replies. This is indeed a lot of food for thought, the trade-off between megazoom and IQ is quite clear to me now, and it looks like I will have to carefully consider which one to prioritise, because, as you say, one cannot find a universal camera that is good in every category.

Looking at 100% crops (which I realise may not be a fair measure), it seemed that FZ150's predecessor, the FZ100, is doing even worse than the SX1/30, so I'm not sure that the FZ150 brings that much improvement, and it seems there aren't too many reviews of it out there yet. In any case, it will probably be considerably more expensive than SX1/SX30 (because it's newer), and I'm not sure it's worth it.

I've finished comparing all the megazooms currently on the market, and one more contendant has come up - the Sony DSC-HX100V, which seems to have slightly better IQ, plus all the rest of the qualities that the other two have. I'll be comparing between these three now, also keeping in mind the dSLR with 3x zoom option.

Thank you once again for your help!

 

[longtalker]

I have one more question about megazooms vs DSLRs. One thing I notice in all megazoom samples, in addition to the noise/lack of detail/watercolour aspect, is quite a high geometric distorsion of all images (mainly barrel distortion), as can be seen, for instance, in this sample. Is this problem also related to the high range of the lens, which means it has to cover too much ground with only one set of optics? Would a typical dSLR 3x lens not suffer from visible barrel distortions like the one in the linked picture?

 

[GardenersHelper]

Looking at 100% crops (which I realise may not be a fair measure), it seemed that FZ150's predecessor, the FZ100, is doing even worse than the SX1/30,

Definitely (IMO).

so I'm not sure that the FZ150 brings that much improvement, and it seems there aren't too many reviews of it out there yet.

The message I'm getting at this early stage (I haven't pored over the samples myself yet, just glanced at one or two and dipped into the discussions at dpreview) is that at base ISO the FZ150 is probably a bit better than the FZ100, and a lot better at higher ISOs. Indeed, it may be better than other recent superzooms at say ISO 1600. All of which may or may not turn out to be true of course, and in any case may or may not be relevant for how you will use your new camera.

In any case, it will probably be considerably more expensive than SX1/SX30 (because it's newer), and I'm not sure it's worth it.

This is true.

Just a thought - have you considered a second hand/refurbished SX10is? I don't know anything about its video, but it meets your other criteria. I know some people think the SX10 was the model that Canon got "best" in terms of how many pixels to stuff on to a small sensor, and that the SX20 and SX30 were slight downgrades in terms of IQ.

I've finished comparing all the megazooms currently on the market, and one more contendant has come up - the Sony DSC-HX100V, which seems to have slightly better IQ, plus all the rest of the qualities that the other two have.

Presumably you are ok with an LCD that tilts on one axis rather than the 3-axis rotation of the SX30 etc?

I'll be comparing between these three now, also keeping in mind the dSLR with 3x zoom option.

If you are considering dSLR, you might want to consider the Panasonic G3. You are presumably aware of the difference between the phase detect focusing that dSLR's use and the contrast-detect focusing used by micro four thirds cameras like the G3 (and superzooms). They have operating characteristics which are (IMO) significantly different, and which is best for you depends to some extent on what types of subject you will be handling and how you like to do your focusing.

One other thing specific to the G3. They have (IMO) messed up the exposure bracketing a bit. With three shots, you can only get from -1 to +1 stops. If you want to go wider you need to use more shots. To get from -2 to +2 you need to use 5 shots, and to get from -3 to +3 you need to use seven shots. Since exposure bracketing is a key factor for you, I imagine this alone might be a killer for the G3. (The SX10, like the SX30, is three shots, up to -2 to +2.)

Oh, and one other thing. If you are considering using the LCD for composition/focusing etc rather than the viewfinder, be sure to try the candidate cameras in a shop. I know that dSLRs have live view these days, but I was really keen on using a Canon dSLR, and I wanted to use live view, but when I tried the 600D and 60D I was really surprised and disappointed - I found the live view implementation very kludgy, and unusable for my purposes. (The G3 is built around live view use and works fine in that respect. The live view on the Sony SLT cameras like the A35, A65 and A77 are probably fine too.)

 

[Nawty]

Is this problem also related to the high range of the lens, which means it has to cover too much ground with only one set of optics?

That's right

Most dSLR lenses also have some levels of distortion and typically speaking the more you as a lens to distortion (and other issues) you get.

 

[arad85]

I have one more question about megazooms vs DSLRs. One thing I notice in all megazoom samples, in addition to the noise/lack of detail/watercolour aspect, is quite a high geometric distorsion of all images (mainly barrel distortion), as can be seen, for instance, in this sample. Is this problem also related to the high range of the lens, which means it has to cover too much ground with only one set of optics? Would a typical dSLR 3x lens not suffer from visible barrel distortions like the one in the linked picture?

The distortion in the image you have linked to there isn't barrel distortion, but pure geometric distortion due to the fact it's a wide angle lens. Barrel distortion is a bulging of lines that should be straight and is typically worse the further out you are from the centre of a lens. The ultimate barrel distorter is a fisheye...

Basically, the wider you go, the more important it is to hold the camera so the plane of the sensor is perfectly vertical. Any deviation from that and lines will do as they do in that picture - go in or out at the top (depending on which way the plane of the sensor is out relative to the subject).

 

[HoppyUK]

I have one more question about megazooms vs DSLRs. One thing I notice in all megazoom samples, in addition to the noise/lack of detail/watercolour aspect, is quite a high geometric distorsion of all images (mainly barrel distortion), as can be seen, for instance, in this sample. Is this problem also related to the high range of the lens, which means it has to cover too much ground with only one set of optics? Would a typical dSLR 3x lens not suffer from visible barrel distortions like the one in the linked picture?

Yes, wide range zooms tend to have more distortion - can go right the way from barrel at one end through to pincushion at the other. However, this is easily corrected in software and many of the better compacts these days put it straight automatically (along with many other image enhancements) in the conversion to JPEG.

But as Arad says, in the image you've linked, that's not distortion in the true sense at all. It's just 'exagerated perspective' as a result of using a wide angle lens, close to the subject, and pointing up (not square to the buildings). The top of the building is obviously much further away, relatively speaking, so it's smaller.

You would get exactly the same result with any lens/camera used in the same way, from the same position. Conversely, if you move back and use a longer lens, the effect is dramatically reduced or even eliminated. You get the opposite effect with longer lenses, compressed or 'flatened perspective'. Both effects are a function of distance, nothing to do with the lens as such.

 

[GardenersHelper]

I have one more question about megazooms vs DSLRs. One thing I notice in all megazoom samples, in addition to the noise/lack of detail/watercolour aspect, is quite a high geometric distorsion of all images (mainly barrel distortion), as can be seen, for instance, in this sample. Is this problem also related to the high range of the lens, which means it has to cover too much ground with only one set of optics? Would a typical dSLR 3x lens not suffer from visible barrel distortions like the one in the linked picture?

FWIW, I think most of what you can see in that image is not barrel distortion. Here is a version of it to which I applied rotation and then perspective correction.

http://www.flickr.com/photos/gardenersassistant/6136861196/
(I will of course take this image down if there is a copyright issue. I'll assume just for now that there isn't.)

The lines that should be straight are now straight I think. (Beware of some optical illusions arising from the white building not having vertical corners above the first floor.) With barrel distortion straight lines would be curved, increasingly towards the edge of the image. I didn't correct for that type of distortion, and I don't see much if anything by way of such curves.

However, the SX10 (and I'm sure other superzooms too to greater and lesser extents) does have some barrel distortion at the wide end. Not a huge amount, and correctable, but it may well be more than you would get with lenses that cover a narrower range of focal lengths. However ....

.... I believe that some cameras (including some dSLRs and micro four thirds cameras) apply corrections in software to adjust images for barrel or pincushion distortion, and possibly also for vignetting. This may result in a corrected geometry being visible in JPEG versions of images, but uncorrected geometry being visible in unprocessed RAW images.

How well such corrections can be done depend on the software knowing the characteristics of the lens in use, which might be the case for (some of) a manufacturer's own lenses, but I imagine not normally for third party lenses. With a superzoom however there is just the one lens, and so such corrections may be applied as a matter of course to superzooms and the distortion (in JPEGs) may be quite mild. I do recall correcting the barrel distortion at wide angle for my SX10, but it always seemed to be fairly mild distortion at worst.

 

[longtalker]

Nick - Many thanks for the corrected version. Yes I would be ok with just a tilting LCD rather than a fully articulated one, if the IQ at low ISOs were visilbly better! The IQ improvement that the Sony brings over the SX30 is minimal though, so the two factors even out. Even the barrel distortion is the same as for the Canon (as per by above post), even though on paper, both models have very low geometric distortions!

I am definitely considering getting either a refurbished or an unboxed/slightly used camera. The SX10 was until recently in my shortlist, because its image is clearly superior to that of its newer siblings, however it does not have HD filming, which I decided is quite important to me.

I can live with just 3 steps of exp bracketing (I've had none so far, and have long thought it would come in handy) :-D

Looking at some of the samples from Panasonic G3 and from Canon EOS 600D, the IQ is indeed beyond reproach, however I notice there is still a tendency for barrel distortion (or rather, geometric distortion, as arad85 explained), if not as pronounced as in the megazooms. I now better understand the reason for that, and appreciate that, in similar shooting conditions (distance to subject and camera angle), even a dSLR image would look the same.

 

[longtalker]

OK so I compared (using DPreview's "Compare Studio Samples" tool) 100%-crop samples from the new Panasonic FZ150 with those from the Sony HX100V, and the former are clearly more crisp (almost as crisp as the unlikely hero which my old Canon A610 has become!), therefore I think I will wait a bit until this camera becomes available in the UK, and if it's not much more expensive than the Sony, I'll probably go with that one. I will give the same grace period to the soon-to-arrive Canon SX40, which I expect will also bring an improvement over the average IQ of current megazooms. If these two new cameras are too expensive, or if they otherwise prove to be disappointing, then I'll go for the Sony, which seems, overall, a better choice for me than the Canon SX1 or SX30.

Thank you very much again for all the helpful inputs of those who've replied, and I'll keep this thread updated in case anyone follows it who has a similar dillema to the one I had.