Discuss theCanon EOS T1i / 500D DSLR Review- tell us what you think of the Canon EOS T1i / 500D.
Printable View
Discuss theCanon EOS T1i / 500D DSLR Review- tell us what you think of the Canon EOS T1i / 500D.
Great review, Bryan!
Just one question, though:
Increasing sensor density lowers the DLA (Diffraction Limited Aperture)
by magnifying the diffraction effect - making it more visible.
This number is the result of a mathematical formula that approximates
the aperture where diffraction visibly starts affecting image sharpness
when viewed at 100% full-size on a display (which is what a high
percentage of site users are doing to evaluate their gear)
I can totally understand that flaws in the optics (like diffraction) are more visible within a 15 MP image than say, a 12 MP image. But, why are do we compare 100% crops from the T1i with say, an 12 MP Xsi image? Shouldn't we observe these crops at 80% with the T1i, and 100% with the Xsi?
The same goes for noise - Shouldn't crops be observed at 80%, and then compared?
Sorry if this stirs debate - I'm sure Daniel Browning will be on this soon[:)]
Thanks for bringing a thoughtful review of the T1i to us so quickly, it was great!
I'm surprised to see the IQ improvement too, it just makes me think "ahh, how nice that would be if my 50D can do the same...". I'm still satisfied with my 50D for what I use it for, and we can expect an even better 60D or whatever in the future with less noise now [:)]
Another excellent review. I'm sort of tempted by the T1i for macro/astro (manual focus uses in which I often can't use a whole full frame field anyway).
What a long way the rebel line has come from my old (and much loved) XT.
Quote:
Originally Posted by alexniedra
I may well be misunderstanding you, alexniedra, but diffraction is not a flaw in optics. It is just a fact of life that the image of a point through a lens is not a point but a small disk (not really a disk but a diffraction pattern, but might as well think of it as a disk). The size of the disk on the ccd is proportional to the f number. That is, double the f number and the diameter of the disk doubles. When the disk is bigger than a pixel, you notice it. When pixels are smaller, you notice the disk sooner, and DLA is lower.
So, if diffraction is a flaw, it is a flaw in the Universe, not in the lens.
As much as I like the review and respect Bryan for his attention to detail, I mislike the following choice of words:
Quote:
Originally Posted by Bryan Carnathan
Perhaps it would be better to say "revealing" instead of "magnifying". Increasing sensor density does not change diffraciton, it only makes it easier to see. I know that seems picky, but there seems to be a misconception about this.
Sorry- Bryan already said he had given more attention to DLA than was warrented, and I agree he covered the subject sufficiently. Yet here I am talking more about it :)
I've been looking to get into photography and have already committed to the EF-S 17-55 2.8 lens, so I've been waiting for quite some time for the reviews to hit for the T1i. Unfortunately, things are at 2 extremes right now.
Bryan, your review paints a pretty good picture of the camera -- good image quality, good noise handling, a durable shutter, and no mention of overly odd white balance or dynamic range issues. The only low point, which is unfortunate but probably not a deal breaker for me, is the slow AF system when taking continuous shots. Movie mode quirks notwithstanding, this seems like an excellent camera for the money. The mobility it offers is attractive as well.
However, another review came online today over at digitalcamerainfo.com. They paint a drastically different picture of the camera, particularly in areas that I care a lot about -- image sharpness,dynamic range, and white balance :(. Their resolution measurements were very, verypoor anddynamic range fell off drastically starting at ISO 400 and above.
They were routinely getting very soft images and things were so bad they asked Canon for a second camera which continued to perform below par. Now they did say they tried multiple lens but I'm not sure which ones they tried.
I still would reallylike the T1i as the 50d is sadly just out of my price range after dropping the cash for the expensive lens. Would you or someone else with a T1i be able to post sample pictures? Also, might it be possible to see how the camera performs on the ISO 12233 Crops when compared to say the 50d when using the same lens?
JesseY, I was a little confused by the digitalcamerainfo.com review. They kept saying "the camera" wasn't sharp, but they're testing it with the kit lens. Do they mean to say the *lens* isn't sharp?
Of course, dynamic range is a different matter. I'm a little surprised by (and suspicious of) their result.
My rule of thumb is, when in doubt- Bryan is right. The other reviewer is wrong. [;)]
Quote:
Originally Posted by JesseY
I would be interested to know at which apertures (f-number) they did their tests. As the T1i has a Diffraction Aperture Limit of f/7.6, testing the camera even at f/8 would reduce sharpness due to diffraction. On a side note : Canon's Digital Photo Professional seems to apply default noise reduction since a few versions. As the sharpness and the contrast both decrease when noise reduction is added, the issue they noticed in their review could come from this.
As a XTi owner, I think I'm gonna pass this one and maybe go with a used 40D when I'm gonna be able to afford it. The DLA is an issue for me, as I do macro shots and fireworks photography, which requires apertures between f/8 and f/11.
I think that it would be a good idea for Canon (and other SLR manufacturers) to stop cramming more pixels in sensors, as even though your picture is bigger, the increased diffraction can be a serious problem if you require a big depth of field. They should try to produce full-frame sensors at a lower cost instead, as the photosites are bigger than a 1.6 crop factor body for the same resolution.
Quote:
Originally Posted by STL
They did their tests at a variety of f numbers. The measurement of sharpness was number lines resolved across the whole picture. Diffraction is not an issue here.
Quote:
Originally Posted by STL
Since they tested lines per whole picture, more pixels could only do better, not worse, than fewer. Diffraction is not changed by increased resolution. Diffraction gets worse only when you stop down.
Quote:
Originally Posted by STL
Macro would be one of the few cases in which I would think the new rebel would be superior to the 40D. It has better low noise performance and better resolution, both of which are important in macro since macro requires lots of light and since one tends to crop macro pictures. You'll be using manual focus most of the time, and probably won't care too much about super fast frame rate or responsiveness (thus much of the advantage of the 40d is negated). The rebel's superior screen will be of use for manual focus with live view.
Hi Everyone,
Alexniedra - The purpose of the 100% crops is to show you exactly what you get from the camera. I can add the XSi results scaled up to the T1i's pixel dimensions if that is wanted. They will be a little softer than the T1i results, but not dramatically different. Scaled down T1i results would be sharper.
Benjamin - Right - What will the 60D bring us? Likely full frame rate 1080p HD video at least.
Jon - Thanks! I'll change that word.
I'm looking to remove ALL confusion over DLA. I first included it to help people understand why their images were soft - to answer the question "Why is my f/22 photo soft even when I used a $2,000 L lens?" (to take the extreme example). I fear I might have still caused misunderstanding for some.
Jesse - That is a great lens - good choice in my opinion.
I'm not seeing any T1i image attributes that noticeably stray from the previous Canon bodies. Canon's auto white balance under tungsten has never been good, but it is easy to change the selected white balance in-camera or in-post. Though I didn't take a scientific measurement, I didn't notice any significant change in dynamic range. You can look for differences in the color block example.
The lens used for the testing can make a huge difference in the results. I have both the 50D and the T1i results loaded for the Canon EF 200mm f/2 L IS USM Lens (link to comparison) in the ISO 12233 Chart tool, but I discovered that one of the 4 flashes was not firing for the T1i test. I'm going to be replacing these results soon.
There are a lot of other factors that can make big difference in image quality. Shooting RAW can be one of them. Noise reduction is another. Sharpening is another (the ISO 12233 chart results have very little sharpening applied). Comparing one brand to other further complicates the process.
STL - While your plan is not a bad one, DLA will probably only be an issue for you if you are reviewing your narrow aperture results at 100% on screen or printing to a huge size. And even so, the T1i will give you more detail to work with. I think a low-cost full frame body would be a big seller. A refreshed 5D (add sensor cleaning, Live View ...) would be very popular.
Regards,
Bryan
I think I'll wait until the next Rebel comes out and get a T1i when it's in the $500.00 range. Then my XSi would be an excellent second body. I like that the batteries are the same on the two models.
Quote:
Originally Posted by Jon Ruyle
I also shoot shows for my university's newspaper. Therefore, I need the 40D's better AF accuracy in low light, as the XTi sometimes hunts for focus in these conditions. I also like the fact that the 40D has 9 cross-type sensors instead of a single one in Rebel bodies. Even though both cameras have only a sensor optimized for low light (the center one), I would like to be able to focus with the other points as well for portraits, etc. without having to think about the orientation of the sensor I want to use.
Note that I did not listed shows in my list of uses, because diffraction is really not an issue at f/1.8 !
Still, I understand that the T1i's better performance in low light is a good thing for shows and the like. I will keep my eyes open for a 60D, even though it might cost twice as much than a used 40D.
For macro shots, I like the fact that the 40D focussing screens can be changed (for the Ef-S screen, namely), which can be useful to preserve battery life (even though it's possible to focus using Live View). I liked the split-screens of film SLRs and this is the closer I can get without going with a third party screen.
On a side note : are SD cards as fast as Compact Flash ? I like the fact that I won't need to buy other memory cards if I buy the 40D over the T1i.
Quote:
Originally Posted by Bryan Carnathan
Touché ! Some people were expecting a split in this model line (3D and 7D) before the 5D mkII was announced. Even though the 5D mkII seems to be a fantastic camera, its price point is way too high for many amateurs like me. A 7D of some sorts (a refreshed 5D) would be a nice addition to Canon's lineup.
Quote:
Originally Posted by STL
Don't hold your breath. As I explained in the following thread, the price premium for full frame is going down very slowly:
Does it make sense? An $1800 Full frame DSLR from Canon
A refreshed 5D would actually have a higher price than the 5D2, because the R&D amortization savings would be far less than the loss in marketability due to missing features. (Just as film cameras that lacked autofocus were more expensive, despite having lower R&D costs, because manufacturing cost savings were very little compared to the loss of market.)
Not meaning to argue like a DPR forum equipment-measurebating fanboi, but I find this section of the 500D review discussing the AF a bit odd:
Quote:
I humored myself by shooting a youth soccer match with the T1i. As expected (based on my experiences with the other Rebel models), I threw away a high-90-some percent of the images of players running toward the camera because they were OOF (Out of Focus). The Canon EF 300mm f/2.8 L IS USM Lens I was using is easily up to this task, but the Rebel AF system, even using the f/2.8 cross-type center point-only, is not. If you are shooting sports with a T1i, setup for shooting stills and action that remains a constant distance from the camera. Better yet, step up to one of the higher Canon EOS model lines.
Isn't the 500D using the same AF module as the 450D, 400D, 20D & 30D? [*-)] Here's the review for the exact-same 9-pt. AF on the other cameras:
450D: The DIGIC III processor gives the XSi / 450D's AF system an approximately 30% speed improvement over the XTi / 400D. In addition, the XSi / 450D is also able to recognize a larger amount of defocus with its slightly longer vertical component of the center focus point.
...
Putting numbers on the realized advantages between the three AF systems I've discussed here is not easy - and it will depend a lot on what the camera is used for. Single Shot AF is working very well and I'm getting many very nice shots in AI Servo mode tracking action. I am certain that I am getting better results with the XSi / 450D than I got from the XTi / 400D... All said, the XSi / 450D gives us a very competent autofocusing system - it is excellent for the price.
400D: The Canon EOS Rebel XTi / 400D has inherited the 30D's high-precision 9-point autofocus unit. A sophisticated dual-precision/cross-type sensor in the center AF position is supported when f/2.8 or wider lenses are used. Focusing performance under low light has been improved by 1 stop (EV -0.5). My experienced focus hit rate has been very good with this lens - including the results from sports action at several soccer matches. The AF improvements are very welcomed.
30D: The Canon EOS 30D shares many of the 20D's excellent specs and features including the same 8.2 MP CMOS sensor, DIGIC II Image processor, 5 fps fast continuous shooting, high-precision 9 point AF (Auto Focus) system, viewfinder and a durable magnesium alloy body. These are the same core components that delivered the excellent performance from and image quality of the EOS 20D.
20D: The Canon EOS 20D seems to have cleared up this issue. Autofocus is both fast and accurate. Although the Canon EOS 1D Mark II is still superior for focus speed, I'm impressed with this upgraded feature on the 20D - it is excellent for a non 1-Series camera body.
...
The 9 focus points employed by the Canon EOS 20D are a nice upgrade from the 10D's 7 focus points, but not close to the 1D Mark II's 45 points. I missed some of the focus points when utilizing AI Servo to track action at a recent soccer match. I sacrificed my framing slightly to get the focus correct. And as I said before, focus accuracy is very good. My hit rate is noticeably higher with the 20D than with the 10D.
***
So my point is, ifthe AFwas good on the semi-pro 20D & 30D & previous Rebels, shouldn't the same module perform the same (at the very least) or better on the 500D due to the newer and faster DiGICIV and (presumably) improvedAFalgorithms? I'm just wondering why it suddenly turned "unusable in sports" when I never heard the same complaints when the 20D & 30D were, in their time, the semi-pro cams that were often used for sports/action by those who can't afford the 1-series. Thanks in advance for anyone who can give an explanation or their experience with these bodies that the same AF module.
Hi Eugene,
I don't like how that sounds (re-reading what I wrote days ago) - too harsh - I'll rewrite that.
I too would like to hear how other people are making out with the T1i's AI Servo performance - I was not very successfulat the game I shot with it. Results were similar to what I got from the XS.
Regards,
Bryan
I understand Bryan, no problem with me, I just found it curious as to why the AI Servo performance wasn't as good as those from the 20D/30D/400D/450D before, when it uses the same AF module. Also, I noticed this in your review:
Quote:
And while it should not be news-worthy, the Canon EOS Rebel T1i / 500D's sensor arrived clean out of the box and has not needed a cleaning to date. I will note that this is not the latest Fluorine-coated system.
It seems to contradict what the Canon USA press release linked at the bottom of the 500D review said:
Quote:
EOS Integrated Cleaning System
With the introduction of the EOS Rebel T1i camera, the entire Canon EOS system is now equipped with the highly acclaimed EOS Integrated Cleaning System. The Self-Cleaning Sensor Unit for the Canon EOS Rebel T1i has been upgraded with a fluorine coating on the low-pass filter for better dust resistance.
Lastly, this one, an issue which continues to confuse me since I keep reading contradictory info on other sites discussing the 500D specs:
Quote:
One of the significant upgrades in the T1i over its predecessor, the Canon EOS Digital Rebel XSi / 450D, is the new, up from 12.2 MP and Canon EOS 50D-matching, 15.1 megapixel sensor (though the microlenses are not gap-less like in the 50D).
The DPR & RG sites still state thatthe 500Dalso has 100%-coverage, gap-less microlenses like the 50D sensor, while still other sites say the opposite. Can you tell us if the info you got was directly from Canon? Thanks again, since it doesn't make economic sense to me for Canon to build yet another 15MP APS-C sensor just for the 500D when they already have one for the 50D. If it's indeed true thatthe sensor does not have gapless microlenses unlike the 50D, how then can we explain the apparently better high ISO performance of the 500D compared to the 50D? Is it all due just totweaked onboard image processing, or lower read noise from having just 2 readout channels & a lower FPS needinglower-clocked signalling? [^o)]
Update:
Chuck Westfall has confirmed that the T1i self-cleaning sensor system does not have a Fluorine-coated low pass filter. The CUSA press release is in error.
Chuck also confirmed that while the T1ishares the same photosites as the 50D, the T1i does not have gapless microlenses. The T1i has two readout channels compared to 4 readout channels on the 50D.
Now we know.
Good to know [:)]
Wow, that's a bit surprising, thanks for the update. So how do we explain the apparently better high-ISO performance of the 500D compared to the 50D, if the former doesn't even have 100%-coverage microlenses on its sensor? A tweaked in-cam image processing pipeline? [^o)] Lower read noise due to only using 2-channel readouts & lower FPS needinglower-clocked signalling?
Eugene - That question remains unanswered by Canon (I asked). [:)]
Quote:
Originally Posted by Eugene Bautista
[8-)]
What is confusing? The posts were pretty straightforward, try reading the full 500D review first to get up to speed. Incidentally, DPR's review of the 500D was controversial, as usual. DPR displayed yet again their penchant to pick on little things if it's a Canon they are reviewing. [^o)]
UnbelievableBryan! , what an incredibly written, precise and accurate synopsis/ review. I am a painter withsignificant 35mm photography experience back in the 70's -80's, however today my experience with digital is limited to point and shoot during vacations and family events. Fortunately, my art career has grown to require excellent documentation of my work some of whichmight possibly be used for short run, small format (less than 18x24 +/-) in house produced, framed color prints. Since I have the lighting and basic photo knowledge, I would like to do the work myself. Budget is a significant issue. My first bare bones choice was the Cannon Rebel XS. Then the XSi, and now I am straining to possibly afford the T1i. The price difference is a strain but if worth it I can find a way to afford it. Please give me any advice to help me which you might have a moment to share.
Sincerely
Parker Nicholls