Comparison: Sigma 400mm f/5.6 HSM vs Sigma 600mm f/8 Mirror

I sold my Sigma 600mm this week, but before I shipped it out I did a head to head comparison between the Sigma 600mm f/8 Mirror (Reflex) and the Sigma 400mm f/5.6 APO Tele-Macro HSM. The result was interesting, but not very surprising.

I tested three lens combinations:

1. Sigma 400mm alone
2. Sigma 400mm with 1.4x Tamron-F teleconverter
3. Sigma 600mm alone (well, using FD-EF adapter without the optical element)

In terms of setup, I shot across my living room and kitchen to a (mostly) flat Costco ad posted on a cupboard door. A single Sunpak 383 at 1/16 power illuminated it to add more light and to avoid any camera shake (the flash effectively makes the shutter speed 1/1000th second or faster). All shots were taken at 1/200 sec, wide open (f/5.6 for the 400mm alone, f/8 for 400mm w/ tele and 600mm), and ISO 100 for f/5.6, ISO 200 for f/8. All shots were manual focus and I took at least five for each, choosing the best to display below to avoid focus issues.

Here are the full frame shots (downsampled, click to see larger... but not full-res):

Overview: Sigma 400mm f/5.6 Telemacro HSM

Overview: Sigma 400mm f/5.6 with 1.4x teleconverter

Overview: Sigma 600mm f/8 Mirror

These shots don't really show anything notable for full frame. Yes, the 600mm has less contrast, but we expect that from a mirror lens. The 600mm seems to have more vignetting too, but again, that's expected.

One thing I noticed was that both lenses performed about the same in terms of focusing -- as in, it was a pain in the a**, but in this static situation I had over 50% accuracy on nailing focus. Obviously, the f/5.6 of the 400mm wide open makes the viewfinder brighter, but in this case I didn't find that a brighter viewfinder made focusing any easier.

Next up, a comparison of 100% crops. All were from the center of the shot. These were converted from RAW with 100% sharpening in Bibble and all other settings off.

100% Crop: Sigma 400mm f/5.6 Telemacro HSM

100% Crop: Sigma 400mm f/5.6 with 1.4x teleconverter

100% Crop: Sigma 600mm f/8 Mirror

As expected, the 400mm alone is the sharpest, followed by the 400mm with TC, and the 600mm decently behind. I'd also say the 400mm alone has the best contrast and color; again, this was expected. So, if you don't need 600mm of reach, the 400mm will definitely be your best option.

But what if you do need 600mm of reach? For instance, if you shoot birds, you pretty much need whatever reach you can get. Which lens combination will give you the best reach if you want to see something really tiny or really far away?

I tested this by taking larger portions of each of the images above and upsampling them to all be at the same resolution as the 600mm image. The image below is the composite of these three images (click to see it at 100% crop):

Comparison: All upsampled to resolution of 600mm
Top: 400mm, Middle: 400mm w/ TC, Bottom: 600mm

This is where things get interesting.

In terms of best reach, the 400mm with 1.4x teleconverter gives the sharpest image. Not by a lot, but you can definitely tell in areas of fine text (like the "invent" in the HP logo). But, on the other hand, the 600mm does slightly outperform the upsampled 400mm. Again, not by a lot, and the 400mm has much better contrast, but I can definitely see more detail on the 600mm image.

Of course, these results are the lower end of 400mm image quality. Although Tamron-Fs aren't that bad, they are a lower-end teleconverter, and a Pro TC should be at least marginally better on a super telephoto. Also, these were shot wide-open; if I could have stopped down the 400mm, it would have sharpened up even more. I'm not exactly sure how much more it would have sharpened up, but I suspect it would be noticeable.

So, in the end, yes, you are much better off with the 400mm lens than the 600mm mirror lens. This was my original hypothesis back when I got the 600mm. Plus, the 400mm lens gives better contrast, bokeh that isn't donut shaped, and autofocus. It seems like the 400mm would be a no-brainer compared to the Sigma 600mm.

Until you look at cost.

The 400mm (without TC) goes for $250+. The 600mm goes for right around $100 if you have to purchase the adapter with it. In fact, mine just sold for $70 including shipping without the adapter. Ultimately, it comes down to how much money you can spend and if you are willing to tolerate the quirks of the 600mm.

Personally, I'm happy to have the 400mm, but the 600mm was a great stepping stone until I could afford it.

Sigma 400mm f/5.6: Adorama and Test Shots

This morning I was already planning out some bad things I was going to say about Adorama in a post today. Specifically, they seemed to forget about mailing it out (I e-mailed them on Wednesday after ordering the lens on Sunday and asked why I did not have a tracking number yet; miraculously I didn't hear anything until it was in the mail the next day, UPS 3-day instead of ground!) and they gave me the wrong information on whether the lens suffered from the Err 99 problem.

Then, I e-mailed them this morning, and my tune changed significantly.

Here's the e-mail trail:

Me:

I received the lens and it is in good shape.

Except -- before bidding, I asked if it was fully compatible with a 20D and you said it was. This was the conversation:

Me: "Does this lens work on a 20D?"
You: "Yes, it is compatible with Canon 20D."
Me: "So, to be completely clear, this lens does not have the Sigma incompatibility that causes Err 99s on a 20D, 30D, etc. It has been rechipped (or didn't need the rechip) and works perfectly on newer cameras?"
You: "It should work fine."

But, I tried it out and it will not stop down, throwing an Err 99 on my 20D whenever I am in a mode other than A or M and whenever I try to set an aperture other than wide open in those modes.

I'm a bit unhappy about this -- not that the problem exists, because I was aware that older Sigma lenses have some incompatibilities, but that I wasn't given a straight answer about the condition of the lens so I could adjust my bidding accordingly. I would expect that a used camera dealer like Adorama would know the status of a lens or at least have the resources to verify if the problem exists. I bid on this lens with the understanding that it was fully compatible with my 20D and I am disappointed that it does not.

Is there something that can be done? What are your return policies? Is there a possibility of a partial refund if I was to keep it and save the return shipping (maybe 10% of the price)?

Them:

Good Afternoon Sean, We apologize for any inconvenience this may caused you. Would you be happy with $50.00 dollars as partial refund? Best regards Louis

Wow.

That was way easier than I thought. Of course, I had them over a barrel and they knew it -- I had clearly asked about the Err 99 problem, they had given me the wrong information, and I had bid more than I normally would have. I just did not expect them to respond within an hour and offer a partial refund (my preferred method to resolve it) for more than I had asked for. More than fair!

As Louis said (if you contact them ask for Louis!), "Our customers are very important to us."

So, yeah, I feel like I got a decent deal on this lens now. And while I thought Adorama was a little slimy before, my tune has totally changed now. Their quick response to this issue was very commendable.

Oh, and for those of you keeping score at home, note the importance of establishing a communication trail and the importance of protecting your rights and recontacting the seller if anything is wrong. EBay really is a buyer's market -- don't let yourself get abused!

* * * * *

Anyway, I was able to get some sample pics from the lens today. I was looking for a bird to shoot, but all I found was a nest!


The macro capabilities of the lens are pretty decent, but the inability to stop down (plus the really long and unwieldy focal length) make it non-ideal.


Still, though, I was pleasantly surprised by the sharpness and contrast. I'd love to be able to stop it down 1/3 or 2/3 of a stop to sharpen it up even more and gain some depth of field, but it really is pretty sharp wide open. The HSM is really fast and snappy and definitely on par with the USM in my Canon 70-200 f/4. Obviously, the thing is a tank and heavy as hell, and the slow aperture (f/5.6) makes it difficult to get good frame rates, but it worked well on the monopod.

Most of these shots were ISO 400 (with some ISO 200) and minimal editing other than a bit of contrast. If you doubt, here are some straight from the camera 100% crops (from RAW, default settings):




In my mind, camera shake and focus is going to effect the sharpness way more than the optimal ability of the lens. For the shots I nailed the focus and didn't have any shake on, the sharpness was quite good and the contrast and color were beautiful.

More on this lens in a couple of weeks when I have a chance to put it head to head vs the Sigma. I also put the Tamron-F 1.4x teleconverter on it and results were pretty good.

If you've got questions, leave a comment and I'll do my best to answer it.

Sigma 400mm arrived...

... and I'm a little bit disappointed.

Nothing wrong with the lens, it just exhibits the Err 99 problem. I was really hoping it wouldn't :) I'm not sure if I'll pursue Adorama for it, since the price I paid is pretty decent overall even if it suffers from the problem.

It is one of those things where I'd be shooting wide open or slightly stopped down anyway, so I'm not really losing much. But the idea that I can't stop down bugs me. Either way, it is a stop-gap lens until I can get a better one, so I should be happy. And for $300, it is a darn good lens (as far as I can tell). I'll evaluate it again in the light of day (although that may not happen for a few days).

Other impressions:

• It is really, really heavy.
• It is very big!
  
• The HSM is pretty nice -- seems to lock on solidly, even indoors, and it moves pretty quickly.
• Solidly built (except the hood) and in good condition.
• Nice case and packing.

Sigma 400mm to arrive today...

... and I'm pretty excited.

Of course, I'm working/transporting kids until 9 pm so I won't get to try it out for a while. I may stop home quickly if it is delivered before I need to pick up the kids (not wild about a $300 lens sitting outside my door for too long). That is, assuming a signature is not required...

A few more Sigma 400mm f/5.6 APO Telemacro HSMs have shown up on eBay, and I'm realizing more and more that the HSM usually does not have the Err99 compatibility problem. I suspect it was released right during the time when Sigma realized the mistake so Sigma rechipped many of them before they left the factory and others were sent back soon after purchase. That makes the HSM significantly more valuable on the Canon mount than the others since it is much more likely to allow full aperture control. In fact, there's an HSM lens currently listing for $455 with a little bit of time left, meaning mine may be worth significantly more than just $300 (if I was willing to sell it again).

There's also another lens that just went up which has two spots of mold or fungus on the front element. While fungus is generally a bad thing, it might be an opportunity to get a decent lens cheaply and then get it cleaned.

The current listings are below (this is the same search I've been using to track Tokina and Sigma lenses)...

Sigma 400mm f/5.6 APO Telemacro HSM Overview

I would love to make this a grand post with lots of pictures and links and great information.

Sadly, I don't have an hour or two to write a blog post at this point. So I'm just going to dash this off and then fill in some blanks later. I figure that is better than to write a post about my decision making on buying a certain lens after I receive it (and my concerns and/or suspicions are faded into history).

A while ago I decided that I wanted a super-telephoto in the 400mm range but compromised with a Sigma 600mm f/8 mirror lens. Since then, while the Sigma has worked pretty well, it has left me wanting a refractive lens so I could actually use more of the shots for stock. Out of boredom (or procrastination) I started monitoring eBay for appropriate lenses.

The first thing I realized was just how many versions there are out there. For instance, these are the Sigma 400mm f/5.6 lenses I know of:

• Non-Apo. Generally not well reviewed, and not very common. <$150 used, but probably not even worth it unless you can get it really cheap. I haven't seen a lot of them in EOS mount. Suffers from Err 99 (more on that later).
  
• APO, but not telemacro. Quite common on eBay, with maybe 6-10 lenses posted a month. Generally seen as a decent performer, but not as good as the telemacro lenses. Easily distinguished by a smaller focus wheel. $150-$250 used. Suffers from Err 99.
  
• APO Telemacro. A little less common on eBay -- maybe 3-5 lenses month. The telemacro lenses are very well reviewed and in some cases they've been given better marks than the Canon 400mm f/5.6 USM. A good lens with good optics, but a little on the heavy side. Easily recognized by the monster focus wheel. Most suffer from Err 99. $200-$300 used.
  
• APO Telemacro HSM. Pretty rare -- I've seen two this month, but previous months I didn't see any. Optically identical to the non-HSM telemacro and physically only differs slightly, but includes the HSM motor (high speed manual?), a USM equivalent. The most recent version, and only released in EOS mount and (some other mount I can't remember). Some suffer from Err99, but many don't. $250-$400 used.
  

I've also been keeping an eye out for the Tokina 400mm f/5.6 AT-X SD but that lens is quite rare and does not show up much. Maybe 1-2 a month if you are lucky (beware the manual focus FD version).

For obvious reasons, I've only been looking at the Canon EOS mount since I want autofocus to work on my camera. The big issue with Canon versions of the Sigma 400mm f/5.6 APO lenses is that most were released before the 20D/350D and newer cameras, when Canon tweaked their lens communication code and caused older Sigma lenses to not work. As a result, most 400mm f/5.6 lenses cause an Err 99 code on newer Canon cameras, locking up the camera.

All is not lost though, since the glitch is something to do with aperture communication when the lens stops down. So, if you keep the lens at f/5.6 (in A or M mode) it does not glitch as much. The down-side of that is you are always shooting wide-open, which makes good lens performance important (which is why the telemacro version is much preferable). Of course, I haven't been able to test one of these lenses, so this is speculation on my part. From a recent eBay auction, the problem is described as:

It works on all Canon EOS cameras, film and digital. However, on a recent digital body like 450D, 30D or the 5D, it must be used at full aperture f/5.6. Auto focus works well and automatic metering can be used in aperture priority mode on these cameras. Manual Focus works always well at any aperture.

The fix for the Err 99 problem is called 're-chipping' the lens, which Sigma used to do for free, but now no longer does it. From what I've heard, 're-chipping' is simply reprogramming the communication chip without even opening the lens up but they call it 're-chipping' out of habit.

Of course, on mounts other than Canon, I've seen no mention of incompatibility problems.

In other auctions, I've seen lenses with the Err 99 incompatibility marked as Not for digital (including the seller adorwin, who I got my lens from, the eBay branch of Adorama). Either way, it is worth asking the seller about the problem. Some have no clue about the glitch though, so beware if you get a strange answer.

In my case, I asked about the problem and got a relatively clear answer but am still not entirely convinced. This was the conversation:

Me: "Does this lens work on a 20D?"
Them: "Yes, it is compatible with Canon 20D."
Me: "So, to be completely clear, this lens does not have the Sigma incompatibility that causes Err 99s on a 20D, 30D, etc.
It has been rechipped (or didn't need the rechip) and works perfectly on newer cameras?"
Them: "It should work fine."

So, I'm 80% sure I'm getting a newer version with no incompatibility. I've definitely heard of the HSM version having the problem and not having the problem. Given the chance of a re-chipped lens, my lens is easily worth the $305 I paid (including shipping). The re-chipped lenses can definitely command a price premium on eBay.

Even without the chip, I was still after a Sigma 400mm f/5.6 telemacro since they are reputed to be quite sharp wide open, and most of the time you'll be shooting wide open with a super-telephoto anyway. It is a nice option to stop down to get a little extra sharpness, but not a deal-breaker.

Watching eBay for a protracted amount of time I noticed a few things that I found odd. For instance, on the lens I won, about half-way through the 10 day auction time two guys got into a bidding war and drove the price over $350. Then, a few days later I checked the price on a whim and found that both bidders were gone! Apparently they both had their bids retracted, which I've never seen before. That seemed a bit odd to me.

Another funny thing was I bid a few times (already in the lead each time) including a final bid raising my max from $283 to $293. Since I wouldn't be around when the auction closed, I wanted to make sure I had the best chance of winning it, but I didn't want to spend more than $300. Well, as it turned out, the guy who was the second place bidder made a final bid of $285. If I hadn't raised my bid by that $10 I would have lost the lens!

Another odd thing I've noticed is the same buyer going after many of these Sigma 400mm f/5.6 lenses even after he's already won one of them. I've also noticed the same seller placing at least two of these lenses up for auction at relatively high prices. I have to wonder if someone is trying to turn a quick buck by flipping these Sigma 400mm lenses. I'll keep watching and try to figure it out!

I have more to say (including the fact that I'm going to sell off a bunch of stuff to pay for this lens), but I'm out of time tonight and I need to get to bed. I should get the lens in the next week or so and I'll at least post an image or two to whet your appetite while I wait until after my defense to get some real time using the lens.

Sigma 400mm Reviews:
Photography Review
Photozone.de

I just won...

... this.

(if you can't see it, it is a used EOS mount Sigma 400mm f/5.6 APO Telemacro HSM)

I got a pretty decent price ($290 + $15 S/H) especially since there is a decent chance it doesn't have the Err 99 incompatibility. Even if it does, it is still a decent price for this lens, since the HSM is quite rare and the condition seems good.

I'll be selling some lenses to cover the cost, including my Sigma 600mm reflex (once I'm satisfied the new lens is better).

More details soon...

THIS is a lens...

I've been randomly searching around for 400mm telephoto primes and/or a decent medium cost macro when I stumbled on the Sigma 200-500mm f/2.8 EX DG APO IF.

All I have to say is... wow. The thing weighs 35 lbs and gives you 500mm @ f/2.8 (or, if you use the included 2x teleconvertor, 400-1000mm @ f/5.6).

Too bad it costs $25K, or the equivalent of a pretty nice car.

I bet it makes some pretty nice images though!

50D vs 5D... or 1Ds Mark I?

Another in a long line of posts that I write, make it a draft so I can do one final edit, then forget about it for over a week.

I guess I am not alone in realizing that the new Canon 5D Mark II opens up some opportunities to snap up the 5D Mark I for a good price. In fact, a lot of people realized that the price of the 5D and 50D are very similar and there's a lot of debate about which is better. As you probably know, I would prefer the 5D.

Bob Atkins (bio), a very knowledgeable person in all things EOS, has an interesting post about the 50D vs 5D debate. Go ahead and read it. Yeah, really, I mean read it. Even if you shoot Nikon or some other camera than Canon, give it a quick read. There's some bigger issues discussed here than just camera models.

Ok, are you back yet?

So, I take no issue with his first point.

In fact, I think it is important for all photographers to realize that diffraction reduces sharpness at smaller apertures. I knew this somewhere in my head, but I also distinctly remember wondering a few months ago why my images at f/22 weren't as sharp as I hoped they'd be. Put simply, after f/16 or so, expect sharpness to decrease as you stop down. This happens for all cameras from all makers. Bob also has a good point about the trade-off between spherical aberration and diffraction and the basic 'sweet spot' of any lens.

On the second point I have to cry 'Bad Science!'.

Yes, I get his point; the 50D's 15.1 MP is better than the 40D's 10.1 MP and this difference shows up even with crappy lenses. If you can call the kit lens 'crappy'. Actually, the kit lens isn't that bad, which many have been saying all along.

My main issue with Bob's comparison is that the experimental evidence suggests that APS-C resolution peaks at around 12 MP (again, see my previous post for details). Honestly, I suspect it might be somewhere between 12 - 15 MP, but the point remains... of course the 10.1 MP sensor in the 40D is outresolved by the 50D!

Also, I have to say the evidence Bob provided was less than compelling -- the JPEG artifacts in those images were more noticeable than the resolution differences. I consider my eyes pretty good, but I was hard-pressed to see a big difference in those images. And for that reason, the other factors, like the cost and improved screen, would be more important to me than the extra resolution when deciding between the 40D and 50D.

Ok, maybe my cry of 'Bad Science!' was a little premature. I do see Bob's point. He's mostly poking holes in other peoples' arguments that the 50D has no increase in resolution over the 50D. And he does demonstrate that there is a meaningful difference in picture quality between the 40D and 50D even on the kit lens. It sounds like he is taking aim at those people who are trying to compare 100% crops between the cameras -- of course the 100% crop of a 15 MP camera will be less 'sharp' than a 100% crop of a 12 MP camera. I fully agree -- pixel peeping is not what makes a camera useful.

I still really wish Bob had posted some better images though. It might even be cool to see the difference on an "L" lens between those two cameras. It'd also be really interesting to see the difference between a 12 MP sensor and the 15 MP sensor, if that's possible.

* * * * *

Ultimately, I think we are splitting hairs on this 40D vs 50D comparison. If you have the cash, buy the 50D. If you don't, buy the 40D. If you are somewhere in between, compare the features (I'd consider the 50D more of a 13 MP camera than a 15 MP though) and make an educated decision.

Then focus on your pictures.

After all, a picture of a turd is still a picture of a turd whether it is captured in 10.1 MP or 15.1 MP!

* * * * *

A little addendum to this post which is (somewhat) related.

Turns out, the 5D Mark I isn't the least expensive full frame camera on the used market.

The Canon 1Ds is.

On eBay, the 11 MP 1Ds prices have dipped below $900 while the 12.8 MP 5D is still scratching $1000. A local craigslist ad had a 1Ds for $950, well below used 5D prices.

Sadly, there's a reason you can pick up a used 1Ds for less than a used 5D. The 1Ds is from the previous generation than the 5D Mark I, meaning worse noise performance and a reputation as a very large, heavy brick. On the plus side though, it is full frame and it is a 'professional' camera, meaning weather sealing and an intimidating presence.

For more, check out:

• Photo.net thread: 100% crops comparing the two cameras. Not a huge difference visible.
• Sportshooter thread: Echos most of what I've seen on the web -- the 1Ds has significantly worse low noise performance, a slow and hard to use interface, and weighs a ton.

If I could get one cheap, I might consider the 1Ds for a stop-gap camera. Plus, I'd look really good with a professional camera around my neck (and permanently adjusting my body to a stooped position).

But I'm still left wanting the 5D to have it both ways: full frame AND low noise performance!

Great Starter dSLR: Canon 20D for $330!

If you missed it (because I typed in the wrong date and buried it), check out my last post, My Next Camera... a 5D?

I was searching through B&H the other day (just gained affiliation with them -- they are an awesome site for new and used equipment, some of the best prices and selection around) and I noticed something...

The used Canon 20D has dropped in price significantly.

This is a big deal. A year ago when I bought my 20D (my review) it was the second best prosumer Canon model available and commanded a price of $600-$650. Now, you can get a used 20D at B&H for $330. That's a significant discount (it is noted as a 'Back to School' special), but it appears the price of 20Ds has plummetted in the past year since the 40D and 50D have entered the market.

(Note: I'm not sure if the link above will work for very long. But, you can check current 20D prices by going to B&H, selecting Used Products in the search type box, and entering Canon 20D in the search area.)

The reason this is so exciting is because the 20D is the perfect digital SLR for entering the Canon line. For $400, you can get a great body with a good starter lens (I recommend a Canon 50mm f/1.8 ($90 new, $60 used) or even the kit lens ($50 used)) which is perfect for a photography student or budding photo enthusiast. This is the best deal I've ever seen on a starter kit, and the real difference (in terms of pictures) between the 20D and top of the line 50D is very, very slim. Yet, the difference between the 20D and 10D (or Digital Rebels) is HUGE: the 20D gives you much better ergonomics than the Digital Rebel series (if you have normal size hands; small handed people might like the Digital Rebel series better), and major improvements over the 10D like instant startup, ability to mount EF-S lenses (a deal-breaker right there, honestly), and better low light performance.

The 20D has the best value of any of the Canon dSLRs. There's no contest.

I kind of wish I'd made this post BEFORE Christmas because I suspect it might have helped out a few parents. But, better late than never.

If you are interested in a used 20D, B&H is a great place to go for used bodies. You also may be able to get a good deal on a body and some lenses on eBay -- here's the current listing of 20D's:

My Next Camera... a 5D?

Welcome to all of you who got cameras for Christmas and are starting to look for accessories!

I have to admit, I'm a bit of a scrooge (grinch?) when it comes to Christmas. I'm not really sure why, but it just seems like the holidays are too much work for the fun -- and that's even though my wife does most of the holiday prep work!

This year it felt like I was the president of the AV club, from twenty+ hours of work on a photography project (more on that later once the gifts go out), kid portrait shoots and editing, and videotaping the kids in the Christmas Eve service (which isn't edited yet!). I don't know why it bothers me so much -- I generally enjoy the work and editing, but people saw me with a tripod and automatically assume I know what I'm doing. At least four people asked me for a copy of the service which is a bit... disconcerting... because my rig (and experience) isn't great and I hope they aren't expecting something great.

Maybe I'm just too much of a perfectionist, but I hate handing out anything that doesn't reach a high standard. Don't get me wrong, I think I did a decent job, but my goal was good home videos, if you know what I mean.

* * * * *

I didn't really get much for Christmas, which is honestly to be expected now that I'm older. I know I'm hard to buy for, and when it comes to things like photography equipment, I'm pretty particular and like to pick things out myself.

Actually, if I'm honest, sometimes I enjoy picking out the equipment more than using the equipment once I get it. There's just something I love about researching camera stuff, savoring the choices, planning how to get a good price, etc.

Yeah, I know it is dumb.

Santa did bring me a new 4GB CF card. I heard he got it for $15 after a rebate. Apparently memory of all types is dropping in price, so it's a good time to buy!

* * * * *

Gear-lust brings me to the point of this post. Actually, this post has been festering (and yes, I really said 'festering') for about three months. It all started when I heard that the new Canon 5D Mark II was slated to come out this fall. Of course, in my gear-lust, my first thought wasn't "Wow, I want to get a 5D Mark II," it was "I bet used 5Ds are going to drop precipitously in price -- I should get one of those!"

Yeah, I'm a little weird. I realize that. But I need to dream about stuff that's attainable. Not dream about a $3K camera that I can't afford.

Like most amateurs (or pros with cashflow problems), I've been planning which camera I'd get next for quite a while. For the longest time I was lusting after the 40D or 50D. Then, I realized that the extra resolution of the 50D was pretty much past the resolving limit of even the best lenses (in a crop sensor, my research tells me you don't really need to go beyond 12 MP before you bang into the limits of lenses -- see DPReview assessment). Sure, I'd still love a 50D (the VGA screen is stunning, I have to say), but I'd probably get a used/refurb 40D if I was to buy right now.

To be realistic, though, there's very little I could do with a 50D that I can't do with my 20D. Ultimately, most of my problems in photography relate to what I put in front of the lens (and how I light that) than anything on the camera.

I try to keep that in mind, but that doesn't quench the desire.

I think photography enthusiasts fall into two camps: those who are in love with the art and those who are in love with the science. I definitely fall into the latter camp.

But, there is a significant difference between a full frame camera like the 5D and APS-C frame cameras like the 50D. A larger image sensor means better wide angle, better resolving power*, and better low-noise performance. I'd like all of those things. In fact, I think all of those things would let me get high quality shots in spots I can't right now.

Good quality using available light indoors. Higher resolution suitable for stock. Experiments in extreme wide angle. Bumping the resolution of all my full frame lenses instantly!

Which is why I'm shooting for a used 5D next. Granted, it will be probably half a year before I can even consider it, but the price should continue to drop while 5Ds are dumped into the used market as the pros and rich amateurs upgrade.

Ken Rockwell predicted this predicted this back in September -- at the time, used 5D bodies were going for about $1,500. Since then I've been tracking the used price of 5Ds on eBay -- these prices represent the lower price-point for working cameras (which assumes you'll spend a little while bidding):

• 9/20/08: $1,400 (new for $1900ish)
• 10/8/08: $1,350
• 11/24/08: $1,200
• 12/26/08: $1,050 (a few went for lower than $1,000, but most are in the $1,100 range)
  

As of this writing, B&H has new 5Ds for only $700 less than the 5D Mark IIs ($2,700 on 12/26/08). That's not a very significant discount. Amazon still has a pretty high price on the 5D Mark IIs, but the 5D is a little lower and a 50D goes for $1,070 on Amazon.

Between a 50D and 5D, I'd go for the 5D almost every time, unless I needed the reach for wildlife or sports.

Of course, I'm not in a position to buy anything right now, and if I was, I'd be buying lenses or accessories because I need them more and they're less expensive. Specifically, I still want a used super-telephoto, a macro lens with autofocus, and a flash that supports E-TTL.

Still, it is nice to dream. And maybe, just maybe, once we have more money I can find a used 5D for a good price.

* * * * *

* Here's a little appendix on the resolving power thing:

From the experimental data, it appears 12 MP is about the limit of even the best lenses for the APS-C form factor. That means it is pointless to go beyond 12 MP -- sure, you have more pixels, but each means less. Even looking at that 12 MP image at 100% resolution, you'll see less detail and sharpness than a 6 MP sensor.

Extending this to full frame (see sensor size comparison; full frame is approximately 2.6 times larger than APS-C) this means the theoretical resolving limit of a full frame camera will be about 31 MP. Actually, I suspect it might be a touch less since most lenses have less resolving power at the edges of the frame. Either way, the 5D's 12 MP is well below that limit and that means each pixel means a lot -- a 5D should give a noticeably sharper image at 100% crop than the 40D even though the 40D has slightly fewer pixels.

The 5D Mark II at 21 MP is nearing that theoretical limit, but I think it is safe to say each of those pixels will mean a lot if you have a good lens. In both lines though, continuing to push the MP would be pretty stupid. I'm hoping camera makers will instead shift to pushing dynamic range, where I think a LOT of improvements can be made by rearranging sensors on the chip and using multiple sensitivity pixels.

Review: La Crosse BC-900 Battery Charger

A month ago I posted about three of the best high end NiMH battery chargers out there and mentioned that I decided to purchase and review the La Cross Technology BC-900 AlphaPower Battery Charger. (Note: the BC-900 has been replaced by the La Crosse Technology BC-9009 AlphaPower Battery Charger, which appears to be the same package, just a different color, and for the same price). While the charger has its user interface quirks, steady use over the past few weeks has proven to me that it is the best battery charger deal out there, period. The link above takes you to Amazon, the cheapest place I've found for the charger (currently $40 shipped).

Before I get into the details about how well the BC-900 works, let me explain why it is a good deal. Actually, the picture above will save me some time -- the $40 package at Amazon contains everything shown above, plus a few things I couldn't fit in!

Specifically, you get:

• BC-900 charger and power supply
• Carrying case
• 4 AA batteries (2600 mAh)
• 4 AAA batteries (1000 mAh)
• 4 C adapters (letting you put AA batteries into something that takes Cs)
• 4 D adapters (requires C adapters to adapt As to Ds)
• Instruction booklets

What makes this such a good deal is the extra batteries (worth $5-8) and the C/D adapters (which I haven't seen anywhere else).


Charger Functionality:

A top of the line battery charger should do three things: charge, discharge, and let you know what is happening. The BC-900 does all these things very well.

Sure, in the instructions and sales literature it talks about four different modes (Charge, Discharge, Test, Refresh) but really, all the modes just boil down to combinations of charging and discharging:

• Charge: Just charge the battery until it is full.
• Discharge: Discharge the battery completely, then charge it until it is full.
• Test: Charge the battery, discharge the battery, then recharge the battery to full.
• Refresh: Discharge the battery fully, then recharge, then discharge, then recharge, etc. Stops when the capacity (measured during discharge) stops increasing. This can take DAYS, but is a good way to rejuvenate really old batteries.

Another really great thing about the charger is the ability to charge (and control the mode) of each battery separately. This is awesome for me, because I no longer have to worry about matching batteries during charging, or discharge batteries to get them to match. 95% of the time I just need to charge batteries at the slowest rate (200 mA) which is the default. So I just pop them in the charger and remove them when they say full, even if the other ones aren't done yet.

I can also control charging rate (200 mA, 500 mA, 1000 mA, 1800 mA (two batteries max)) with a few pushes of a button. Discharge rate will always be half of charging rate. If you do the math, this means that the test and refresh modes take a long, long time to complete at the lowest charging rate. Since most newer batteries don't suffer from memory effect (inability to take full charge if you don't discharge them before charging) most of the time I don't bother with the discharge cycle. As my batteries age, I'll probably start doing full discharges before every fifth charge or something like that, but for now, I'm not too worried.


Display and User Interface:

One of the great things about the BC-900 is the display. In addition to displaying the mode, it also displays the instantaneous voltage (good for estimating how much charge the battery has or how soon it will finish), time spent (dis)charging, capacity in mAh, and (dis)charging current.

Specially notable is the capacity display shown in mAh. For instance, in charging mode, it will display how much charge the battery has taken. This lets you know how much the battery was used or if the battery is starting to suffer from memory effect. Similarly, the discharge mode displays the capacity discharged or charged depending on the current state. Finally, the test and refresh capacity shows the status of the most recent discharge (for the most accurate estimate of capacity). Not too many chargers let you know exactly how much capacity an old battery has left; the BC-900 does. That alone is worth a lot.

Also, there are buttons which let you set each battery mode and charging rate individually, but I have yet to master them. The first few times I tried to set battery rates/modes individually I reset the modes of the other batteries (not a big deal, but I need to spend more time figuring it out). Again, though, most of the time I just need to charge at 200 mA, which is as simple as putting the batteries in the charger and walking away.

I did use the 1000 mA charging rate once when I left an SB-20 on all night and rendered a set of batteries (that I needed) stone dead. So, I popped them in at 1000 mA for a half-hour while I was setting up other things and had enough charge to get what I wanted done.

One thing that really bugs me is the lack of a backlight or LEDs. Just a single red/green LED for each battery would let me see at a glance when the batteries were finished charging without crossing the room. Minor, but on my other chargers, I find it very handy to be able to see if they are full at a glance.


Batteries and Adapters:

While batteries are pretty cheap commodities (good NiMH batteries cost between $1.50 to $2.00 each) it is always handy getting extras. In this case, the batteries alone are worth maybe $10, easily boosting the value of the BC-900 package over the BC-700 charger (which is near identical). And what can I say, the included batteries are... functional batteries. The first few charges didn't get me to full capacity, but now I'm getting 2400-2500 mAh out of them.

A big plus for me is the adapters included in the package. Yes, they are pretty cheap and no, I haven't needed them yet. But, since similar adapters cost about $2 each new, that's another $8 added to the value of this package. Again, this is an EXCELLENT deal.

Add the carrying case in, which while not necessary, is useful, and it's an even better deal. The carrying case, by the way, has slots for the adapters and would be useful for carrying around a store of batteries with the charger. Why I'd need to do that... well, maybe if I was shooting with three-four speedlights on a regular basis, it'd be handy. I have no problem taking it for free though!


Verdict:

I'm a numbers guy -- I love numbers and stats. And more than anything, this charger gives me numbers. I can tell exactly how much charge a battery has taken, check the status of my batteries, or just pop batteries in whenever I need to and know that they'll be charged safely.

And that is the real key -- convenience. The La Crosse BC-900 makes battery charging very convenient for me and gives me all the information I need to maintain the large stock of batteries I need around my house. If you use a lot of batteries (which is pretty much any family) you WILL save money with this charger.

Speaking of money, honestly, I would have bought this charger for $40 without the extras. Add the extras in, and I think this is one of the best purchases I've made. I know I'm starting to sound like a broken record, but there's a reason I keep repeating this stuff.

I love this little charger!

Note: On Nov 2nd 2009, I started this post and the charger was $38. On Nov 4th, the price bumped up $2 to $40. On November 7th, the price dropped $3 to $37 (with the tag 'Friday Sale'). Not sure why these fluctuations occur, but I'd recommend monitoring the price for a few days if you have the time. Don't wait too long though, with the newer model out (La Crosse Technology BC700 Alpha Power Battery Charger) at only a few dollars less but no extras, I suspect they may phase out the BC900 soon. Current price is below: (link removed)

Note: Almost a year later, it appears that the BC-900 was replaced by the BC-1000. Consensus is that it is the same charger, maybe with a few bugs out of the firmware, and a different color to the case. I highly recommend the BC-1000, especially with the package (batteries and such). My charger (and batteries) is still working perfectly! The updated link to the Amazon page is below:


Lighting Setup:
For those who enjoy lighting setups, here is the setup for the battery shoot. White seamless in my garage, charger/subject sitting on a cake pan (don't tell my wife!). Key light was an SB-20 shooting through an umbrella quite close to the subject at camera left (to keep spill off the background. To further reduce spill on the background, I put the umbrella cover over the back side of the umbrella.


Background light (SB-20) through a red gel. In the image it is positioned in front camera right with a clothes gift box lid as a gobo to keep the red off the subject. Blue hair light which cross lights with the key at minimum power (1/16th). Looking at this image I think I should have restricted it a little bit to keep it off the background, but that caused only minimal problems in post.

The trickiest part of the shoot was getting the key at the right angle to illuminate the display. I'm still disappointed that I couldn't keep the display numbers from shadowing on the back of the display.

Sonia Optical Trigger Problems

I've got a bunch of random things to post today, so stay tuned.

First up, an update to my Sonia Optical Trigger Review.

After plenty of time in the garage studio, I'm realizing my Sonia trigger doesn't seem to be syncing like I think it should. It works with direct, hard flash or large, very bright sources (like bounced from a pure white background), but it has trouble with diffused or small sources over distances of 4+ feet. This is quite a surprise to me since no one else has reported these problems, and I wonder if my peanut slave is faulty.

The problems were highlighted when I was trying to sync from my on-camera DIY ring flash (Sunpak 383) to a Chinese optical trigger (with eBay radio trigger on top) to a radio-triggered shoot-through umbrella to the Sonia trigger on an SB-20 as a background light. I know this sounds a little Rube Goldbergian but in theory it should work well. In practice, the ring flash triggers the radio trigger and the umbrella well, but the Sonia slave doesn't trigger. It DOES trigger if I manually fire the umbrella or if my ring flash is closer to the receiver (3 feet instead of 5 or 6). It is almost like multiple sources are causing the Sonia to not trigger.

Either way, I'm spending too much time messing with optical sync and not enough time shooting, so I'm definitely going to do something to address it. I'll post more when I learn more.

Update: After a little more experimentation, the problem is definitely from the multi-level triggering scheme I was using. I switched the the radio transmitter on the camera (instead of relaying it through the other optical trigger) and suddenly I was getting perfect triggering. So, somehow relaying the signal made things fail. I've got an idea of why (I suspect the slight delay from the optical trigger caused the Sonia trigger to lose edge detection) but for now I just need to avoid that situation and use the radio triggers to trip the flashes that feed the optical triggers. In other words, maybe my Sonia trigger isn't defective after all...

Review: Sonia Optical Slave

As I mentioned last week, I love my Nikon SB-20s but I could not get them to reliably sync with my cheap Chinese optical trigger from eBay. So I went ahead and bought an alternate optical slave trigger from an Indian company named Sonia and put it through its paces this week.

My one-sentence review:

The Sonia slave syncs great with the SB-20, but it's still an inexpensive optical trigger and has limitations.

Hardware and Build Quality:

Most eBay Sonia slaves have a female PC-sync and are packaged as a pair with a hot shoe -- at least, that was what I purchased. Sonia slaves are available individually with other connection mechanisms (FlashZebra has a range with male PC, female PC, microphone jacks). While I'd love if they'd packaged the male PC-sync slave (so I could slap it directly on an SB-20 if I desired) I expect it was probably a safety-type issue to make sure the multiple-sync capability worked.

Yes, I said multiple sync. But more on that later.

I went with the hot shoe option because the price was about the same as the peanut alone and I need a 1/4" screw mount to fit my flashes to my tripod/light stand/monopod. For the record, I paid $12.95 + $2.01 shipping, which is relatively cheap for a slave + hot-shoe mount, but still the cost of half an SB20.

Actually, maybe I should refer to everything in terms of SB-20s now... I get paid about one SB-20 an hour at my job and my rent is a little under 100 SB-20s a month! Yes, I know my hourly rate is too low compared to my rent -- I live in the bay area after all! Anyway, back on topic...

First, the slave itself.

There really isn't much to it -- a few components soldered together and epoxied into one solid little package about the size of my thumbprint. By inspection, it looks like a few resisters, a cap, a transistor, and a phototransistor at the business end. My phototransistor isn't centered very well which may be influencing my sensitivity (or maybe that's by design). The female PC connection is solid and unlikely to break (good thing too -- I could never fix it if it did). I'd have no qualms about letting these little guys float around in a gear bag -- they'd never break. FYI, all pictures are clickable if you want to see the units larger.

The hot shoe is definitely a big step up from the Chinese optical triggers. Instead of plastic, the entire unit is made from metal. I suspect it is machined because I see no obvious mold marks, which is remarkable for such an inexpensive unit. It really is a tank and you'll break your hot shoe before you break the foot of this thing. The back is marked with "Sonia, multi-terminal slavettach solid state" in retro RCA fonts. Somebody in India likes their antique radios. And you gotta love the solid state to let us know they didn't cram any vacuum tubes inside!

The most important part of any hot shoe is the connections, and there's no disappointment here. The peanut attaches to the shoe very securely. All three PC connections (two female, one male) are a step above what you usually get in a cheap unit. I don't foresee anything breaking in the future. And the hotshoe on the top is simply a tank. Overall, the whole thing is the most solid accessory I've seen. As you'd expect, it has some mass to it, but it's small enough that it doesn't really matter.

Funny thing actually -- the other day I found that the back of my Chinese optical slave had fallen off. Apparently it had gotten hot enough in the garage to make the adhesive fail and the back just popped off on its own -- luckily it wasn't structural and I snapped it back on easily.


Functionality:

Of course, what really counts is how an optical slave trigger works. And in that respect, I can't tell it apart from my Chinese optical slave other than the fact it works with a Nikon SB20. That's not necessarily a good thing.

Indoors, it works pretty well as long as there is a direct line of sight from another flash to the eye, or at least a line of sight from a bright reflection (like white paper) to the eye. The only time I've had trouble getting sync was when the trigger was in shadow, as you'd expect. Once you get things set up, the triggers work as their supposed to near 100% of the time.

Outside, though, it is a whole different story. In deep shade with the source flash on high, close, and directly illuminating the trigger you may have a chance. In the sun, or near sunlight, the Sonia unit will just not trigger. When I was shooting the bees I tried a lot of different approaches (including shading the sensor) and nothing worked. So with cheap triggers, the verdict is still OUTDOORS = BAD, INDOORS = GOOD.

I should also mention that there are reports of the Sonia slaves not working with the Canon 580 EX II. I believe I read something somewhere about a mod to the Chinese slaves to make them work on the 580 EX, but it isn't really possible on a Sonia since everything is encased in epoxy. Anyway, you've been warned.


Multiple Sync:

If the solid build of the hot shoe unit wasn't enough to motivate you to pay the extra few bucks for it, they've also added a multiple-sync capability with 'built-in diodes'. What this means is, if you attach the peanut slave to the male sync socket, you can attach a flash to a hot shoe and two other flashes to the female PC sockets on the sides with PC cables. And yes, everything will trigger at the same time!

It turns out I have PC cables for my Sunpak 383 and another hot shoe around the house (I had totally forgotten about them), so I verified multi-sync capability with all three flashes. Of course, I couldn't get a picture because my 20D refused to omit the preflash.

So, the image at left only shows two flashes going off. But I was seeing all three go when I fired my 20D.

I love the idea of using this little baby to triple the power of the flash (or cut the recovery time). Sadly, the best place to use that feature would be outdoors, where the trigger refuses to even go. But, I may try adapting a radio trigger to female PC-sync to make a super-powerful three-flash rig.

While multi-flash capability isn't really a killer app inside, I'd rather have the capability than not. And if I had a longer PC-sync cable, it'd let me add even more to a studio rig.



Conclusions:

If you use SB-20s or might buys some in the future, you need to buy this optical trigger, period. Chinese triggers just don't cut it with those units.

If you don't use SB-20s, I still might recommend the Sonia triggers because of their superior build quality and multi-sync capability. I'm not sure I'd pay double for Sonia triggers, but a few bucks is definitely worth it.

If I get another flash, I'll likely pick up another Sonia trigger and/or another wireless trigger. Either way, I'm definitely not buying any more of the cheap Chinese triggers.

If you're curious, here are the current Sonia trigger listings on eBay. eBay seems to have the best prices for the triggers, but that could change -- make sure you look around!






Note: I recently posted an update about some problems I've been having with the trigger.

Some SB-20 Sync Problems

I set up a little studio in my garage (details forthcoming) and in doing so I learned a few more things about the SB-20 that I'd like to share. Namely, it is a little quirky when it comes to syncing, and I think I've finally figured out some of the problems.


The Multiple Trigger Problem:

First, I've known for a while that if an SB-20 gets triggered twice in a row (rapidly) it seems to dump all the capacitor charge into the tube on the second trigger. In other words, if you have it set to 1/4 power and hit the button twice very quickly, the first flash will be 1/4 power and the second will be whatever is left (3/4 power).

In practice, this is less of a problem than it sounds, because even when my 20D is at full speed (5 fps) it isn't fast enough to cause a problem. But I noticed the effect a lot when I had a crappy connection between the SB-20 and my eBay wireless trigger. To solve the problem I rewired the connection (clipping the end of the trigger and attaching a separate 1/8" plug) and it has been very reliable ever since.

In other words, don't worry about it, but understand that a poor connection or rapid cycle time can cause the SB-20 to become inconsistent.


The Chinese Optical Slave Problem:

This second problem is a bit more serious and I had to find alternate hardware to get around it.

Apparently, the cheap Chinese optical slave triggers (seen in the picture above) lock the SB-20 up after each fire. The slave will fire the SB-20 the first time, but then it won't respond. Removing the slave from the flash will reset it as will pushing the test button on the back of the SB-20. Actually, the first time you push the test button on the SB-20 there isn't a flash, but then afterward, since it has reset, it will flash each time you push it.

Once I figured this out, using all three lights (I only have two radio triggers) meant firing a single shot, manually pushing the test button on the SB-20, shooting again, etc. In other words, it very much interrupted my flow.

FYI, I verified the problem on both SB-20s yet my eBay trigger works fine with both of them. So it is definitely some sort of incompatibility between the SB-20 and this specific model of optical slave. Sadly, this unit (a boxy little guy with a hemispherical clear window and a PC connection) is the primary optical slave available at the low end of the market.

But, there's another slave out there made by Sonia, an Indian company. So I ordered one of them (including a hot-shoe mount with two extra PC connections) and I really hope it won't cause the same problems on my SB-20s. If it works, I'll probably get one or two more to have the option.

If you are interested, the current eBay listings for optical slaves are below. The sonia slaves have a clear epoxy package with some sort of color to them (from the circuit board inside). The Chinese slaves are the boxy ones. I'll review the Sonia slave once I receive it!

Auto Mode Review: Sunpak 383 vs Nikon SB-20

If you are poor like me and can't afford the newest (newer) flashes ($200+ for a 430EX) you have to make due with less expensive or older flashes. In my case, I went with the Sunpak Auto 383 Super (review) and multiple Nikon SB-20s (review). While I was able to save 50-70% of the price of a 430EX on each, I had to give up some features.

These older flashes do not support the newest flash balancing algorithms like E-TTL II (for Canon -- Nikon has an equivalent). E-TTL II is pretty neat because it meters the flash through the lens, exposing based on what the camera sees, and also takes into consideration the focal distance of the lens (full description here at Bob Atkins' site). Of course, neither my SB-20 nor my 383 supports TTL (through the lens metering) of any sort with my 20D. Instead, I must either balance the flashes using manual controls (awfully slow and fiddly) or use the auto mode.

The auto mode is an automatic metering mode based entirely on the flash. First, the user sets the flash to auto and the camera aperture to the value specified for the given auto mode (often, flashes have a few options for auto mode for different apertures). As the flash fires a photocell on the flash records how much light is reflected back from the subject, and when a predetermined amount of light is recorded, the flash is quenched (shut off). Essentially, this lets the user get automatic exposure without any communication between the flash and the camera.

Both the 383 and SB20 support auto mode. The 383 has a photocell just above the hot shoe (the green circle on the left of the image):
And the SB-20 has a photocell in the lower right of the large red window on the front of the flash:
Auto mode is very easy to implement because it is directly on the flash but has a few limitations because there is no camera-flash communication during the exposure. First, since the photocell can't see exactly what the camera sees through the lens, the results will be inaccurate for very wide or very telephoto lenses. Also, since the photocell is at a different level than the lens, it could be occluded by objects or not aimed at the same place when in close. Finally, since the photocell measures all the light coming in, you'll get a bad exposure if something bright is closer to the camera than your subject.

That said, if you just want to get some snapshots and don't want to worry about balancing the flash in manual mode, auto mode is very useful. Put another way, auto mode lets the flash automatically adjust for distance -- in manual mode, every time you move closer or farther away from the subject, you'd need to adjust the exposure. For kids, that kind of constant adjustment isn't possible!

When I got my Sunpak 383, I quickly realized the auto mode wasn't very good. I just couldn't seem to get a good exposure! After that, I kind of abandoned auto mode until the other day when I thought I'd try an SB-20 in auto mode and did a little comparison. Turns out my 383 just has a crappy auto mode (not sure if I can generalize it to all 383s though -- does anyone else have a 383 that works well?).

To start out, I tried my SB-20 in auto mode by switching the main switch to A, set the ISO 100 with the slider at left, and set the desired f-stop using the slider at right (F/8 in this case -- note the yellow indicator). This means I should be able to set my camera to F/8 and get good exposure no matter what my distance to the subject is. The back of the SB-20 looked like this:

So, I fired off the following shot of the stuff sitting on top of my television (all of these shots are pretty much straight from the camera):

SB-20 (auto f/8) @ 17mm, f/8

Chimping the result, I thought it looked a little hot, so I dialed it down one stop, and got:

SB-20 (auto f/8) @ 17mm, f/11

Then, stepping back and zooming in to test if the auto mode was working, I got:

SB-20 (auto f/8) @ 50mm, f/11

So, by the looks of it, the auto mode on the SB-20 works very well. And, in hindsight, I really didn't need to stop down below the SB-20's settings -- it was very close to good exposure and the stopped down version is a little too dark. Since I usually underexpose flash shots a little bit (you can always increase exposure from the RAW file, but if you blow whites out, you can't get that detail back), I'll probably stop down 1/3 of a stop compared to the flash setting, but it really seems like the SB-20's auto mode is right on.

Now, on to the Sunpak Auto 383 Super. Auto settings on the Sunpak are a bit more confusing. First, I set the left-hand slider switch to the red A setting and the top slider to ISO 100. In the F/stop window, f/5.6 shows up (actually, more like f/6.3) which is what I should set my camera to. Also, I set the bottom slider to full (A) -- my understanding is that the auto only works when the slider is set to full (disclosure: I faked the f/stop value because it was in shadow, but that's how it looks!).

Note that the 383 has three auto settings, but as far as I can tell, it is only a mechanical calculator. Going from red to yellow doubles the distance and opens up two stops (consistent with the inverse square law: doubling the distance should require four times as much light).

So, given these settings, I set my camera to f/6.3 and took a shot under the same conditions as above, and got:

SB-20 (auto f/6.3) @ 17mm, f/6.3

Obviously, way too bright and I completely lost all the highlights on the fan. So, after some playing, I settled on f/11 as a similar exposure to the SB-20 (yes, that's 1 2/3 stops down!):

SB-20 (auto f/6.3) @ 17mm, f/11

SB-20 (auto f/6.3) @ 50mm, f/11

So, obviously, the SB-383 works, but it needs to be stopped way down. Consistency (in this tiny test) seems to be good across images, but it disturbs me that the Sunpak runs about 2 stops too hot.

Another thing that worries me about the Sunpak's auto mode is the photocell is very, very narrow (less than a millimeter in diameter, about three millimeters deep). I wonder if it gets very good coverage of a wide angle lens. Even more problematic is the height of the photocell on the 383 -- with a hood on a lens (or even a relatively large lens) the cell will be occluded for at least part of the range. I feel like the SB-20 has a much better design.

My conclusions? I like the SB-20 auto mode a lot better, and it just reinforces that if you are on a really tight budget and need a flash, start with an SB-20! For just $30 you get a flash with zoom, bounce, manual mode, and auto modes. The only thing the Sunpak has that SB-20 doesn't is a rectangular head (useful for putting modifiers on it), a little more power (GN120 vs GN100), and a swivel head (so you can bounce of the ceiling in a portrait orientation. While each of those is significant, paying three times as much tempers the advantages. There's a reason I have three SB-20s and may buy up one or two more...

Anyway, if you'd like to get an SB-20, check eBay. Prices seem to be pretty stable at $30 (including shipping) but there are often even better deals if you look hard enough. Here's the current listings if you are curious: