Sigma Vs 550ex Vs 580ex. A personal Saga. (Beware long post!)

Discussion in 'Canon EOS' started by ben_rubinstein___manchester_uk, Nov 22, 2004.

  1. I?m a wedding/event photographer shooting Canon. I grew up with the
    AE-1 and A1 but graduated to AF Canon as, to be honest, my MF just
    isn?t good enough for wedding photography.

    I graduated to digital recently, running it alongside film with a
    D60. I sold it pretty soon after I had bought it, the AF was
    inexcusably slow and the flash exposure sucked. A couple of months
    ago I sold it and traded up to a 10D. The 20D came out that week and
    the money I saved on the newer body helped to facilitate a 17-40L and
    70-200L to go with it. I have had 30X20" prints made from the 10D
    that equaled my Mamiya 645 for quality.

    I used to shoot weddings and events with 3 sigma ef-500 super
    flashes, one on camera and two on stands using wireless ETTL.
    Portrait shooting was done using the built in light slaves and
    metering the output as my 2 EOS 50 (Elan II) cameras could not do
    wireless ETTL with ratios. I used to manage a photo lab, I know what
    good negs look like, and the flash exposures using that setup were
    spot on when used with moderate FEC for white dresses etc.

    With the 10D and the sigmas I realized that I had to work out a
    system to obtain good exposures. Before I start, in a post below when
    I complained about the inadequacy of Canon's ETTL, I was recommended
    to use CF4-1, manually select focusing points to ensure it was on the
    faces, trade up the 20D and change my sigmas for canon flashes.
    As far as the first two suggestions are concerned, although, yes it
    would give me more reliable exposure, I strongly object to having to
    change to a completely different style of shooting, and especially
    one which I could not cope with when faced with very fast moving
    situations which need split second focusing and shooting, such as PJ
    wedding photography, just to cope with Canon's inadequate flash

    I worked out a system using the sigma flash, that a permanent FEC of -
    1/2 stop would in most cases keep the highlights from blowing across
    all the ISO?s, even when using focus/recompose. However, and I only
    worked this out last night, when the flip down diffuser or an
    omnibounce is used on the flash then the camera may overexpose by 2
    stops or more! I used the flip down almost always as it stopped the
    flash head zooming which could be very audibly whiny during a
    ceremony. With the zoom set manually to 28mm I can still get good
    exposures plus no noise. However I can still not use an omnibounce.
    Bounce flash for some reason is still spot on.

    Oh, another very important point, forget wireless ETTL using the
    sigma flash on digital. If the master can detect no slave, i.e. it is
    out of range, then it will blow on full power regardless. With a
    slave the exposure is wildly erratic to say the least. Please note,
    this is the case with 2 or 3 sigmas, with a canon and sigma (the
    canon as master or slave) or with a canon and 2 sigmas, in other
    words, if you want wireless ETTL with a 10D steer well clear of
    sigma! (One of them was still under warranty, sigma claimed to have
    fixed it but I see no improvement whatsoever).

    I went today to my local pro store to see whether a 550ex was the
    answer to my problems. Terry, G-d bless him, was one of the guys
    working there, and he was wearing a black fleece. In other words, the
    exposure was pretty much the same as if it had been a groom wearing a
    suit/tux. I focused on his face and recomposed so that the focus
    point was on his jacket. I shot an entire 1gig card of Terry with the
    sigma, 550ex and 580ex, using different angles and zooming in/out but
    keeping the focus point on his jacket. I could have set up a more
    detailed test, but I don?t shoot dummies at weddings, I needed to
    know how it would handle real life shooting.

    I quickly established that I could forget wireless ETTL period
    whenever a sigma was part of the combination. I also verified that
    when the sigma was diffused that it overexposed badly, however
    without it the exposures were definitely good enough at a -1/2 stop
    (I shoot RAW).

    The 550ex did not have this problem with the diffuser at all, however
    the exposures were little different to that of the sigma.

    The surprise of the evening was that the exposures of the 580ex on
    the 10D (i.e. no ETTL II) were spot on in almost every frame. (it's
    also a damn nice flash, and finally as small as Nikon's latest the SB-
    800, the feel of it is also a lot better) without any FEC at all.

    I cannot afford to keep trading up my equipment, I'm trying very hard
    to make a living, and as I'm in the lower segment of the market I
    cannot just write off such updates to ?expenses?. I?m sure that a
    20D with a 580ex would make me a very happy photographer, but at
    present I cannot justify the costs, when I am making my sigmas work
    albeit with a struggle. I could sell two of them and buy a 580ex, but
    then I lose my slave flashes that are so easy and quick to set up,
    and that I do my portraiture with when it's too much of a shlep to
    put up my Bowens. I could buy a 580ex and still have no wireless
    ETTL, and that?s probably what I will eventually do.

    Bottom line, firstly congratulations in hanging in there till now! If
    you are going digital, forget the Sigma EF-500 Super flash, the 550ex
    is not that much better, but with the price dropping on it, I believe
    that being able to use an omnibounce and have wireless ETTL, to be
    well worth the extra, especially if you do not have an investment in
    sigma flashes. However, even for the 10D, if you can get the 580ex, I
    really believe that it a big leap as far as ergonomics, weight but
    most importantly exposure is concerned.
  2. So what is your current method of coping with all of the flashes--FEC settings and such? Just curious.
  3. I cannot comment on the Sigma flashes but one complaint I don't understand about using Canon Speedlites is when someone comments that the 580EX provides better exposures than the 550EX. That's because with the E-TTL system, the camera determines the flash power and duration based on the Speedlite's preflash (and the subject's reflectance). Unless the Speedlite is miscalibrated, the 550EX and 580EX should provide identical exposures, unless the camera or flash are miscalibrated.

    I have 3 550EXs and they all provide ideal exposures with my EOS-1v. I know there are flash problems with the EOS digital SLRs and I don't have any answers for that. I am awaiting delivery of a 580EX, which I am buying for its improved ergonomics and additional features, and I fully expect it will perform as well as my 550EXs. Most importantly, I don't expect any difference in exposure.
  4. Don't blame me, the information was there on the histogram, I don't understand it either, but then I don't understand how the 10D would tell the sigma to overexpose so badly when the flash was diffused, when with the 550 and 580ex it didn't...

    I shot around 177 frames with those 3 flashes, it wasn't just a couple with each, and only the 580ex provided consistently good exposures with a tricky subject (white wall, black shirt) without any FEC at all. Both the sigma and the 550 needed up to a stop of minus FEC otherwise the exposure was far too 'hot', and even if the highlights were not blown out, once I had darkened the face sufficiently the background would have gone black, indicating overexposure.
  5. I should have kept the pics but I was reading the histogram for exposure information, not the picture on the screen, and at the time hadn't thought of posting here. I had a job after that and so I formatted the card, sorry!
  6. Fascinating stuff. I too am surprised regarding your comments regarding the 550EX and 580EX. Also surprised that the sigma was consistently goofed up the mix when shooting in a wireless configuration.
  7. The exposure is calculated by the camera, not the flash. But, there *could* be a 'possible' explanation in the zooming VS crop factor, which the 550EX can't calculate like the 580.

    Having said that I must say my 550s work perfectly with my 10D, on location and/or studio.
  8. I was told that the 550EX has compatability issues with the new 20D...I was told the buy the 420EX or 580EX instead. (yes, the 420 over the 550)...I took the 420EX as they had no 580s in stock. I shot a parade last Friday night....and the damn thing would not pick up the flash compensation I was putting into the camera. It was returned on Saturday because they had the 580 in stock. I shot *another* parade Sunday night with the flash. It worked flawlessly!

    That is the single worst thing about going Digital...everything changes at a much quicker pace. So the best thing today is not tomorrow...or worst, doesn't even function! You would think they would build these items with flash-able bios's or they could upgrade them for the newest camera's and stuff...but it's more $$$ for them to force you to buy new gear.

    To me, Digital gear is time stamped...and the value of your gear is squat after a few years. Gone are the days when high end camera gear holds it's value over a long time. Even lenses have computer chips in them and they are tied to certain cameras (if you want their BEST reuslt) So I'm sure the 580EX won't work that well with the new 40D in 2'll have to buy the new 610EX to get all of the features and performance.

    Makes it much harder to swallow the high price tag when I know it will be worth the same thing as my 5 yeard old computer in 2 years...

  9. But, there *could* be a 'possible' explanation in the zooming VS crop factor, which the 550EX can't calculate like the 580
    This isn't the difference in this case, since the 10D is not compatible with this scheme; only the 580 in combination with a newer body (20D and 1D II) will zoom to the right position, and even in that case, it shouldn't make a difference to flash exposures unless you're trying to use the most power the flash can give you. The pre-flash and the main flash are both done with the zoom head in the same position, so the two should be consistent.
    Ben, as others have commented, I don't doubt that the results you got were the results you got :) It's just illogical, given the way flash metering is implemented, and we're trying to figure out how it happened.
  10. I can't explain it either, the differences were better by about half a stop which isn't much! however, half a stop of overexposure on digital is too much!

    I don't know. I may go for it anyway, I was very impressed with the size, ergonomics and recycle time, apart from the exposure. That aside, given how long the 550ex held the fort as the top canon flash gun, the 580ex should be more future proof as an investment than my 10D or even the 20D.
  11. I wonder how a 580 with a 420 slave(s) would work out?
  12. A dumb thought: Could the 550EX and 580EX have a different flash color temperature?

    Could that be affecting the ETTL preflash exposure calculations, and creating the 1/2 stop metering difference?

    Personally, I have not really had exposure issues with my 550EX on my 10D. I have recently been using a 420EX as a slave - and have not noticed any dramatic metering issues (yet).
  13. Ben

    I feel your pain... ;-)

    I have also struggled with Sigma EF 500 Super. I have one Sigma and two 550EXs. When I tried the Sigma as a slave with 550EX as another slave (I used the ST-E2 as the master) I noticed it overexposes perhaps 1-2 stops compared to 550EX.

    I then tried to adjust the FEC on the flash (Sigma) in negative direction to compensate just it. To my surprice I noticed it overexposed even more. Then I tried the unnatural, turned positive FEC on the Sigma flash and when I set it to max positive the results were pretty much the same as with 550EX as slave. I just took couple test shots that time so I have not enough experience with it for strong opinions whether it would work fine in most cases or not.

    Other bad features with Sigma are that if you change the mode (or replace the batteries) it will loose channel and group settings which is sometimes a pain. Also you can't use it as a manual mode slave with ST-E2 because the Sigma has no way to set it in manual mode when it is an E-TTL slave.

    The newer Sigma EF 500 Super DG (or whatever) might be better but I haven't even tried it.

    I think the 10D plus 550EX has a tendency to underexpose which is why I mostly use FEL. Post-adjusting (RAW) images might give good enough results so using FEL might not be necessary.

  14. Oh yes, about the 550EX vs 580EX difference.

    I think Canon improved the accuracy of flash output power control in 580EX. It might explain something.

  15. I have been using a 550 with my 20D now for a few days and not had any problems. It is worlds better exposure than the same flash on the 1D which is horrid.
  16. I agree with Vesa.

    Under the old TTL system, the camera sensor directly measures the amount of light put out by the flash and reaching the film. The camera can signal the flash to stop when when it determines that enough light has been produced.

    With E-TTL, the measurement is done with the preflash which is 1/32 of the real flash output. The camera calculates the amount of light needed and sends the information to the flash. I have no idea what information is sent but it could be something as simple as a multiple of the preflash power (i.e. 16x for half power and 32x for full power). The camera no longer has a way of directly measuring the actual output so tt's up to the flash to accurately produce the requested power.

    If the flash power control is bad, then then exposure will be bad. If the power control is very good, then the exposure should be right on. The secret is to make the power tolerance match the response of the sensor/film.
  17. This may be a stupid idea but since ETTL-II can use the distance, the pre-flash output may be also different. For example, when I used my Metz 54 MZ-3 for macro, the results were horrible overexposure. I just suppose that 1/32 of full power was too much for the sensors at such short distance and therefore the camera did not know how powerful flash is needed.

    If ETTL-II knows in advance that the distance to subject is 2m then it may use less powerfull pre-flash. If the distance is over several meters or if bounce flash is used then stronger pre-flash may be used.

    just idea :)

Share This Page