Future Sensor Sizes - Canon's Westfall...

Discussion in 'Canon EOS' started by eos 10 fan, Dec 4, 2006.

  1. The December 2006 Tech Tips by Chuck Westfall has been posted on DigitalJournalist.org and Westfall answers this Q: "At the present time Canon has 35mm full-frame sensors, 1.3x sensors, and 1.6x sensors, a lineup of three different sizes. End users have various complaints about this, such as lenses behaving differently depending on the body, and difficulty switching between high-end and low-end bodies. Do you have any plans to consolidate your sensor offerings into either 1 or 2 sizes?"
    Read the answer here:
    http://www.digitaljournalist.org/issue0612/tech-tips.html
    --
     
  2. I talked with Chuck myself about this back in November.

    He said that he still saw value in the 1.3x sensor size. However I don't think he's in a position to say what Canon's plans actually are and I'm sure he's not the one making the backroom decisions. All that goes in in Japan, not in the US.

    I didn't take his answer as suggesting either that the 1.3x sensor size would be dropped or that it would be kept. I saw it as kind of neutral, but that he currently saw the 1.3x sensor as certainly viable.

    My own view is that I think the 1.3x sensor is probably on the way out unless it's really needed for speed issues. The major problem with 1.3x is that EF-S lenses won't fit it and EF lenses suffer a 1.3x multiplier, making the widest available zoom (the 16-35) produce an FOV equal to that of a 21-46 on full frame.

    Canon can clearly make a 12MP full frame sensor camera and sell it for close to $2000 without losing money (I'd assume they are still making a profit from the EOS 5D with $600 rebate). How much could they save by putting a 1.3x sensor in it? If people are waiting, thay are waiting for a sub-$2000 full frame camera, not a sub-$2000 1.3x camera. I suppose if Canon could make a 1.3x body for $1500 they might sell some, but I think a 1x body at $2000 would sell better, the though being that sooner or later the 1.3x format would be dropped.

    If you never need a wideangle zoom that goes past 21mm, the 1.3x format does offer the advantages of using the "sweet spot" of full frame lenses and giving you a modest telephoto advantage. There are just no wide, fast lenses. The 35/1.4 becomes a 46/1.4 and the 50/1.2 becomes a 65/1.2 in terms of FOV.

    By the end of February, we should have a much better idea of Canon's plans.
     
  3. I strongly sympathize with the End users have various complaints about this, such as lenses behaving differently depending on the body statement inasmuch as having to see things differently.

    I shot on slide film for umpteen years on my faithful EOS 5 before investing in the Canon EOS 10D. The first lens that went onto the new body was my favourite EF 20-35. Considering all the years I'd been shooting with that lens, I did tend to 'see the wider picture' better. Went out of the window with the 10D! Had to learn a whole new way of seeing! And did I miss my wide perspective!

    Invested in a 5D a while back as a second body, plugged in the 20-35 with great anticipation. Boy, did I rejoice to 'see' the lens work as I liked it to before the advent of digital. Till I went out and shot, that is...

    Having not shot at the widest end for a while (the 20mm was effectively 32mm on the 10D), I'd stopped visualizing things in terms of the 20mm perspective! End result - I needed to start [learning to] seeing things again as my 20mm would.

    So yes, as far as I see it, the biggest issue of lens behaving differently is the perspective... ... at least that's my perspective ;)

    Neville Bulsara
    Travel & Documentary Photography
     
  4. The 1.3x sensor is the only one in jeopardy, if at all. Canon has just released its fastest, most expensive EF-S zoom lens, so the EF-S mount -- with its attendant 1.6x sensor -- surely has a long life ahead.
     
  5. I agree... the 1.3 sensor is not long for this world....it will boil down to two sensors IMHO: 1.6 and the full frame. Because Canon could create a fullframe that can be "cropped down" for speedier frames per second, it seems no longer a need for a 1.3 sensor.

    I would like to see a future 1D series that merges the D and DS lines into one body, and using a fullframe sensor. The 1.6 sensors for the consumer market.
     
  6. I do not see the death of the 1.6x sensor at all. At one point it looked like maybe it would end up in only the Rebel series, but now I think that there will always (i.e. for the foreseeable future) be both a low end consumer 1.6x DSLR (the Rebels) probably priced around $600-$700 and a mid range "prosumer" 1.6x DSLR (the 20D/30D/40D...) probably priced around $1000-$1200

    At some point I'd expect a sub-$2000 full frame DSLR ("son of 5D") and either one or two 1D series models. If there are two one would be optimized for speed of operation and one would be optimized for image quality. The ideal might be to borrow a page from Nikon's book and make a full frame high pixel count DSLR that could also be operated in a reduced frame size at higher speed by only using a cnetral subset of pixels. However that might put the price too high, so seperate models may be the better way to go.

    For sports shooters who need high frame rates and shoot telephoto lenses the 1.3x sensor is no problem at all, it's maybe even an advantage over full frame, so it may well live on in that application for a while.
     
  7. Good post so far. I too invested on the EF-S format and hope it will stick around for awhile. Until competitors offer a full line of full-frame sensors, I don't think Canon feels any pressure to drop a very successful crop size.
     
  8. Canon was dictated to in the matter of a 10mp Rebel. The same may occur again in regard to sensor size.
     
  9. By who? Nobody else even has a full frame sensor, even on their high end DSLRs.

    APS-C will be the default standard for consumer DSLRs, just as 35mm was the default standard for film. Full frame will serve the same function as medium format did in film. Something professionals and aspiring amateurs either need or want.

    Quite honestly, given the image quality of APS-C format DSLRs (at least Canon APS-C DSLRs...), neither consumers nor the vast majority of advanced amateurs actually need full frame, and I'd guess that APS-C quality hasn't peaked yet either.
     
  10. My speculative notion is that before long we'll see something like the Rebel XTi at the high
    end of the 1.6 crop sensor camera line, and we'll see cameras "above" that level using full
    frame sensors.

    In other words, the niche currently filled by the crop sensor 30D would either go away or be
    replaced by a full frame sensor camera at a price that will start out somewhat higher than the
    30D (but lower than the current 5D price) and gradually drop.
     
  11. Dan >> I don't quite agree with you on that one. I see it almost as the other way around. I can see a use for a pro level APS-C camera for those pros that never need to shoot wide,... just long. Some sports photographers, Wildlife photographers, Bird photographers....etc. For some pros, the need to get higher FPS is more important than wide angle.

    Bob >> You are forgetting about Kodak's full frame DSLR.
     
  12. I thought Kodak's FF cameras went out of production.

    I just wanted to point out Canon could take a loss on each 5D sold under the rebate, and easily make it back in the lenses. That's how Microsoft makes money on the XBox, and Apple makes money on the iPod. They loose money on the initial sale, and make it back on software and accessories.
     
  13. I doubt that Jose....most buyers of the 5D already have a decent stable of EOS lenses...I doubt most 5D buyers are first time EOS buyers.
     
  14. Most people seem to think that 1.6 size sensors are comprimise only with no benefits. However, for macro work, the smaller sensor allows for increased magnification and/or deeper depth of field than what is possible with a full frame sensor.

    When I am shooting macro at 3x life size, I need every millimeter of DoF I can get. I don't know if I would choose a full frame camera given the equivalence in color, noise, and megapixels.
     
  15. I think recent advances in fab yield have changed the big picture we were looking at 2 years ago.

    The following sensor cost data comes from EJ Peiker.
    He is a photographer and yield evaluation engineer in the
    semiconductor industry.

    Camera Sensor--------Area (mm2)-----Yield/200mm wafer*---Sensor Cost*

    Full Frame EOS 5D ----- 864 ---------- 2.6 ---------------- $385

    1.3x Crop EOS 1D ----- 511 ---------- 13.0 ---------------- $77

    1.6x Crop EOS 30D ----- 338 ---------- 29.0 --------------- $34

    * Values are an estimate based on semiconductor industry standard costs as of Sept. 2006

    The cost for full frame sensors has dropped dramatically in 2 years!

    Canon could easily put a 1.3X sensor in the next 40D at a price point of $1200. A 12MP 1.3X sensor based on 20D pixel spacing would out perform all current 10MP cameras by a wide margin. The question is will the marketing people let them. Only if the 5D group doesn't feel
    threatened. This would probably require making the 5Dmk2 16MP. Stayed tuned, we will all know in 3 months or so. :)
     
  16. I just wonder now that P & S are putting in APSc size 10meg cmos sensors, and useing good glass such as carl zeiss 24 120 walkabout size lenses, having no dust problems and rivalling IQ of DSLR`s, will Canon & Nikon etc have to bring forward a better set of packages.?
     
  17. Weiyang Liu wrote: "Dan >> I don't quite agree with you on that one. I see it almost as the
    other way around. I can see a use for a pro level APS-C camera for those pros that never
    need to shoot wide,... just long. Some sports photographers, Wildlife photographers, Bird
    photographers....etc. For some pros, the need to get higher FPS is more important than
    wide angle."

    I understand that it is easier to fill an APS-C sensor with the image projected by a shorter
    lens than what would be necessary on full-frame, but I suspect that the market remaining
    for professional crop sensor bodies will diminish greatly once full frame sensor bodies are
    available at lower price points - to the point that it probably would not make economic
    sense for Canon and other manufacturers to build a more "pro" crop sensor camera than
    the ones they currently sell. Just can't see that happening, though I do sympathize with
    those who would like such a critter.

    In the worst case, folks who value the small sensor for this sort of work my just decide to
    put a good lens on something like the 400D/XTi or equivalent, perhaps as a second body.

    Still speculating... :)
     
  18. the smaller sensor allows for increased magnification
    I know what you mean but this statement is incorrect. 'Cropped' sensors don't change the magnification of a given lens (or lens + tubes + close-up lenses). You simply 'see' less of the image circle. Your subject occupies more of the image area, but that's simply because the image area is small compared to a full-frame sensor.
     
  19. Miles Hecker wrote:
    "I think recent advances in fab yield have changed the big picture we were looking at 2 years ago. The following sensor cost data comes from EJ Peiker."
    IIRC, Canon is using 300mm wafers now. Not sure how that would change Peiker's numbers though.
    --
     
  20. I don't see any reason to do away with either 1.6 or 1.3 sensors. Canon will likely want to go down a route of having more choice which means keeping all the sensors sizes alive.

    1.3 sensors are popular with sports and photojournalists which make up a big portion of the professional market. 1.6 will remain popluar in the rebel line, but I also see scope to have a both a full frame 40D and a 1.6 40D in the future.

    It is not really about what Canon wants to do, it is how they percieve they can sell more cameras, and to me this means offering more choice. They produce over 50 lenses, surely half a dozen different bodies is not overkill.
     
  21. I'm not forgetting Kodak's attempt at FF. Everyone else did though and they dropped it from production. It had lots of flaws.

    I don't think it's a question of cost anymore. Huge numbers of consumers are quite happy with tiny 5mm x 7mm sensors. APS-C is huge in comparison, so if they were happy with 5x7, they should be absolutely ecstatic with APS-C.

    There's simply no need for full frame at the consumer or prosumer level, and given the popularity of APS-C coverage lenses there are good reasons for the camera makers to stick with it.

    If Canon made the EOS 40D full frame, or even 1.3x, I wouldn't upgrade from my 20D. If they make it a higher pixel count 1.6x, maybe with increased dynamic range and other bells and whistles, there's a very good chance I will. I assume I'm not alone in this.

    Canon's best bet would be a full frame DSLR at $2199.99 and a 1.6x EOS 40D priced at maybe $1299.99.
     
  22. I wonder if the 30D replacement would go 1.3 or not. Somehow I suspect we'll see a bump up in rez using 1.6. Didn't someone say 1.6 is good up to 12 or 14 megapixels? All I can say is this coming February will be very interesting. I wonder if the 1DsMk3 will go up to 22/ 24 megapixels as some suggest. I'd love to have a 24 megapixels camera so I can get full rez 300dpi prints for 16x20s.
     
  23. If Canon bumped the 40D to 1.3x, few people with EF-S lenses would upgrade. It would be a dumb move at this point I think. There's a LOT more mileage yet to be gained from APS-C. Only when it finally taps out (at maybe 12/14 MP?) will there be any marketing pressure to make the sensor larger and dump the EF-S lens users. By the time the 1.6x sensor tops out, the price on FF sensors may be low enough to use them in mass market cameras if the demand is there.
     
  24. Bob, no longer the P & S are limited,to 5x7 with the APSC 10meg sensor now used, eg: Sony R1.will that place any pressure on the DSLR market?
     
  25. >> For sports shooters who need high frame rates and shoot telephoto lenses the 1.3x sensor is no problem at all, it's maybe even an advantage over full frame, so it may well live on in that application for a while.


    My thoughts exactly. And as there are plenty of sports shooters in the world, it is likely it will stay for some time, at least until Canon can achieve both high FPS and FF.

    Happy shooting,
    Yakim.
     
  26. There is a symmetry to the Canon offerings. The 1DsMkII to the 5D and the 1DMkII to the 30D. The first pair is about max image quality, the second pair is more about speed. The pairs are just separated by cost.

    The replacement to the 1DsMKII will probably be 22+MP. Why take all that sensor for high-speed/sports photos? The 1Ds and 1D are targeted at vastly different markets; as my friend says, "Horses for courses".

    If anything I would see an even bigger split between the 1Ds and 1D lines. I wonder if sports shooters would accept the light loss from a pellicle mirror for 12-15 shots per second and the ability to see the shots?

    Canon seems to like the 20D/30D line at about $1300, leaving plenty of room for a FF camera at about $2000. A 40D at 12MP and the 6D (?) at 16-18MP would be a really nice mix, and bracket the Nikon 200D really well.
     
  27. "Because Canon could create a fullframe that can be "cropped down" for speedier frames per second, it seems no longer a need for a 1.3 sensor."

    Yes they could, and it would be a good engineering compromise.

    I understand some (one?) Nikon camera has the capability to use a smaller area to get speed up - so there may be patent issues which in the world of consumer electronics is a lot more important than good engineering or what the customer may want.
     
  28. Reasons for pro body crop camera..? Just compare the cost, size and weight of a 300 f/2.8 vs. a 500 f/4L. If I were a sports shooter.. would I rather pay for and carry the 500 f/4L to use on a FF camera, when I could do with the 300mm on an APS-C body? A 10-12mp 1.6 crop "Pro" body would be I think ideal for these shooters.. put a better viewfinder, better AF and solid construction/weather sealing all together.
     
  29. "There's simply no need for full frame at the consumer or prosumer level"
    What about the millions of lenses already owned across the world? Not everyone wants to constantly buy new lenses for a specific body.
    As already mentioned, the change in expected FOV of our trusted old lenses is a right-royal PITA. My spectacular Zeiss 21mm became a slightly mundane 33mm on the 20D, but now restored to its intended FOV with the 5D.
    My 85mm f1.2 never got the same look on a crop sensor because the whole limited DOF effect was reduced by the required change in perspective.
    There are of course advantages for telephoto work with a crop sensor, and my 30D comes in very handy for bird photography, but I do need the option of full frame, and I'm sure hundreds of thousand of advanced amateurs would relish a cheap FF DSLR. It also means more high quality pixels can be fitted on the sensor.
    I'm sure the huge P&S market couldn't give a rat's about sensor size, but don't underestimate the sophistication of the world's "on -a-budget" photography enthusiasts who are keen enough to choose SLR's.
    There is really only one argument in favour of crop-sensors, and that is money. Lower cost to the consumer, and higher profit for the manufacturer. As development costs drop, so will the crop-sensors.
     
  30. FWIW, the rumours were that Canon was having difficulty getting reasonable yields at FF with a new technology sensor, and that is what caused delay to the 1Ds replacement. I'd guess that a new technology sensor (?high dynamic range or "Foveon like") could make its entry in a 1.3 crop body as a compromise solution. It is pretty obvious that Canon has managed to improve FF yields for the "old technology" 5D significantly - there is a lot of supporting evidence in Canon's results:

    http://www.canon.com/ir/results/2006/rslt2006q3e.pdf

    I'd expect that given the interval since the DIGIC chip came out for the current 1 series DSLRs that development would now allow at least a doubling of data throughput rates for image processing - the speed constraint if any is going to be on readout from the sensor and will probably be addressed by doubling the number of parallel readouts.

    That Canon has talked about multiple sensor sizes has had some speculating that Canon might even launch a new system to compete with MF backs by offering a larger sensor. Plausible (especially to anyone who has actually looked at a comparison between the 1Ds Mk II and almost any of the current MF backs for studio work), but believe it when you see it.

    As to other uses for 1.3 crop, I think that while Chuck Westfall doesn't set Canon policy, he is more in the loop than outsiders and is always careful not to make comments that are inconsistent with Canon policy, and he is also aware of the feedback to HQ which I think he still handles for the US. 1D series shooters seem to have been able to get away with not having anything realistic wider than the 16-35 unless they opt to use the Sigma 12-24, and to handle the odd effective zoom ranges the crop factor produces - those cameras have become the PJ's workhorses. I really wouldn't be at all surprised if Canon continue to offer 1.3 crop even when the 1D MkIIN is superceded, and quite possibly it may appear at a different price point in the lineup.
     
  31. At least John C. said, this ,"A 10-12mp 1.6 crop 'Pro' body would be I think ideal for these shooters...
    put a better viewfinder, better AF and solid construction/weather sealing all together."

    HOWEVER APS-C and Canon and quality viewfinder is a nonstarter.

    Once you use an EOS with a full frame viewfinder, be it film or digital, it is very hard to go back to that 10D, 20D, 30D (and impossible to a dReb).

    And I more mean that it is hard to justify to buy an APS-C replacement body rather than wait for the (hoped for) FF sensor DSLR announcement early next year.
     
  32. I prefer both the 1.3x and 1.6x to FF. Why? Because I get good results from the "belly" of the "L" zoom lenses with less distortion. The 24-105 is quite good at either 1.3x or 1.6x. I don't know that I would feel as good about it with FF. The 70-200 f/2.8 & f/4 might be OK but I'd rather stay with 1.6 & 1.3. I do enough long lens work to be a very big fan of the 1.6x.

    I hope Canon does upgrade the resolution of the 1.6 body but still keeps the 1.3x.
     
  33. Bob, no longer the P & S are limited,to 5x7 with the APSC 10meg sensor now used, eg: Sony R1.will that place any pressure on the DSLR market?

    I don't think so. The P&S market is dominated by small cameras and you can't have a small camera with a big sensor. You can't place the lens as close to the sensor as you could with film so making small cameras with large sensors is very difficult. You also need a physically larger lens with a larger sensor.

    The argument that people already have 35mm lenses and so want a full frame sensor holds some water, but it's not a deal breaker. 35mm lenses are already usable on all APS-C DSLRs. If you have a 16-35/2.8, it works just fine on your APS-C DSLR. You may need to buy one "superwide" zoom, but that's it unless you chose to do otherwise. I have the same lenses for my 20D as I had for my EOS-3, except I bought a 10-22 lens for the wide end.
     
  34. Bob Atkins: "Canon's best bet would be a full frame DSLR at $2199.99 and a 1.6x EOS 40D priced at maybe $1299.99."

    I hope so. I'm very much ready to move from 1.6x to FF for my non-telephoto work (I'll keep my 20D for the long stuff). Since 24mm on FF is a wider FOV than 17mm on 1.6x, I'm wondering if I'll even use my 17-40 anymore after I buy a FF body. Add the latest features of the 30D (3/5 fps) and 400D (bright LCD, sensor cleaning) to the 5D successor (plus DiGiC II and pop-up flash) for this price, and I'm there with my wallet open!

    Lester Wareham: "I understand some (one?) Nikon camera has the capability to use a smaller area to get speed up - so there may be patent issues which in the world of consumer electronics is a lot more important than good engineering or what the customer may want."

    As long as Canon can prove that they do (essentially) the same thing using a demonstrably different methodology, there will be no patent issue: Patents are awarded for methods, processes, techniques; not ideas or capabilities.
     
  35. To me the set-up I'd like to eventually have is my 20D with my 70-200/2.8IS and a 1.4TC at times; a FF EOS with my 17-40. Then a digital rangefinder with the Eq. of 35, 50 and 90mm lenses.

    Right now I have the 20D set-up, but have to use my EOS630 for really wide, and a Leica CL for my rangefinder and scan the film.

    Funny, I passed over Bob's thoughts on prices before I posted mine. I think those are the right price points, but then again, I thought the 70-200/4L IS would be $800 bucks.

    Anyway.

    The features are more important than the price. If it is to high, it will come down with rebates or the used market. Remote flash control through the built-in flash, auto ISO mode, better AF.
     
  36. To me a 40D at $1299 would be uncompetitive with the latest offerings from Nikon, Sony and Pentax. I know there are good reasons to be in the Canon system, but I think a $1299 40D would be lagging the opposition.
     
  37. There are two ways to compete. One is to cram more pixels onto current sensors. The other is to make larger sensors. The Nikon and Sony lens lines are predominantly full frame. How long do you expect them to stay that way, or why haven't they changed? Maybe they will change after they have mined-out APS-C. The company that has the most to gain and the least to lose by going to a larger sensor because of a unique niche is Fuji. Sigma could gain credibility by going larger than 1.7X. We have 1.7x, 1.6x, 1.5x, and 1.3x, so why not 1.4x and 1.2x?
     
  38. also 2.0x on the 4/3rds cameras...
     
  39. Full frame at a $1000 please.. so us poor people can enjoy in on all the fun :)

    actually what is this new magic chip that is even better than the canon 5d that they are
    having problems with getting yields on? All the pictures i've seeing on the net of the 5d are
    positively unreal.
     
  40. Maybe the 500 dollar full frame digital will be like the Mustang or Camaro with 500 cid motor that passes all smog laws and gets 50mpg. ?
     

Share This Page