Shutter lag of my E-500

[OlyFlyer]

I have checked what many people are complaining about. The way I did may not be flawless, but it gives me some indications anyway. Here are my figures:

Time is measured by waiting until pointer is at 0 then shoot at least 10 pics with each settings. ISO 1600 (don't care about noise), flash turned off.

Manual focus, release button halfway pressed: 0.05s (yes, 50 millisecond). I guess here it mostly depends on reaction speed. Focusing time is not checked.

Autofocus: first point and focus at an object about 5 meters from the cam, then point at the screen and shoot when pointer passes 0. This gave me 0.25s. Lens used is the 14-45mm.

First shot from powered off camera with autofocus gave me 3.15s. Camera is set to reset lens, so focus was far out every time I powered on.

I think these values definitely beat many cameras and proofs that the Oly is definitely OK even for paparazzis, especially compared to non-digital cameras. It is mostly up to your reaction time when it comes to first case anyway. Lens autofocus speed is important for the second case. The third case is the case when the photographer has the camera in the bag. Turn the camera on while lifting the camera if in hurry. This would take quite long time anyway with any camera, especially with the lens protection on. This way the supersonic cleaner has time to clean on the way up to your eyes. So, for all those people who complain about the shutter lag, well, don't complain. Fast CF, good reaction time, and if you are prepared, you won't miss the pic of your life.


This is the site to visit if you want to repeat my not so scientific tests.

http://www.shooting-digital.com/columns/schwartz/shutter_release_test/default.asp

 

[OlyFlyer]

I did some new and shocking measurments. This time using a sound recording software and a sensitive mike very near the shutter button. I set the camera to manual focus, time: 1/4000, flash off, aperture manual. I shot about 100 times pressing the button all the way at once. Each time I had about 250ms delay between the pressing and the begining of shutter opening. That makes my previous measurments just about worthless. I could never go below 240ms. I could really feel that the time between pressing and the shutter sound was like forever.

Anybody has any idea why? What is delaying the shutter in this case?

 

[abhivachan]

When I was reading the book "Shutterfly guide to great digital photos by Jeffrey Housenbold and Dave Johnson". The author mentioned the below points for minimize the shutter lag.

1. Prefocus your picture - save valuable milliseconds
2. Presetting the camera's white balance based on the scene
3. Shoot predictively

 

[OlyFlyer]

Well, I could not agree more. That is also one reason why I redone the test with an other measurment methode. The only thing I missed is to not use Auto WB, but to fix it. Everything else was manual and fixed. That is because I am new to digital photography and on film camera there is no WB. But, could it be that the WB takes such a long time to be calculated? I mean 250ms is a awful lot of time for any microcontroller, CPU, computer, graphic processor or whatever is doing the calculations in my camera.

If it is the WB than it is just an other reason to take all pics in RAW and set fixed WB and fix WB errors at computer time. I have to measure again to see that, but I vill be surprised if it is the WB. I'll be back.

Thanks for the comment.

 

[bliytz]

You would also have to consider that electronics take some time to register button presses.

Old mechanical cameras mostly use the force of the finger to trigger the shutter release mechanism, and so disregards the need to have <half button press detected>-<full button press detected>-<full button press registered>-<calculate auto values>-<initiate shutter mechanism>-SNAP

;p

 

[OlyFlyer]

You would also have to consider that electronics take some time to register button presses.

Old mechanical cameras mostly use the force of the finger to trigger the shutter release mechanism, and so disregards the need to have <half button press detected>-<full button press detected>-<full button press registered>-<calculate auto values>-<initiate shutter mechanism>-SNAP

;p

If there is one thing I know a whole lot about is electronics, microcontrollers and microcomputers. That is the exact reason why I can not find any good explanation about that 250ms delay. That is an awful long time in electronics, even considering some calculations has to be done. But, when there is no calculation to be done since the camera is in all manual, then the time needed to detect the button press should be in the magnitude of microseconds not milliseconds. Half press has no function when everything in manual mode, so the existence of any processing delay in that case is just a design flaw. Just a few days ago I found some discussions elsewhere in the cyber space regarding this mystical 250ms delay, stating it exists even in E-330, which is Oly last baby. Nobody has an acceptable explanation for this. In fact, the delay is the same with manual WB also, so the processing time of that is not the problem. Exactly as in E-330, when everything in manual mode, if you half press the button and measure the time to shutter opening, then it is only a few milliseconds. I am going to contact Olympus to get a more scientific explanation. Hopefully it will result in new firmware release.

Thanks for your comment, I thougt nobody is interested in this subject any more.

 

[bliytz]

That is a wise choice. Do follow up with a post here when you do get information - I am curious too.

 

[OlyFlyer]

That is a wise choice. Do follow up with a post here when you do get information - I am curious too.

I definitly will make it public if I get an answer.

 

[abhivachan]

Please post the same question http://www.fourthirdsphoto.com in technical discussion. I hope, you may get good answer.

 

[VR Man]

Hi OlyFlyer,

Does the shutter depressed signal have to go to CPU of the digicam for processing? If so do you think this will result in a 250ms delay?

Just guessing :dunno:

 

[mpenza]

maybe it's a purposely built-in delay to ensure that everything is ready before the shutter opens? even if you fixed all the parameters, the body still needs to inform the lens what aperture is to be used and to close down if necessary (please correct me if I'm wrong).

Another website does its own testing of such lags - have to say the E500 isn't very fast among the current DSLRs:
http://www.imaging-resource.com/PRODS/EV500/EV500A9.HTM

 

[mpenza]

Updated a chart which I put up earlier for reference (based on tests done at www.imaging-resource.com): E-1, the Olympus first and only "pro" DSLR, is quite a bit faster than the E-500.

camera (prefocus lag/manual lag)
Nikon D50 (113ms/134ms)
Nikon D70 (124ms/155-168ms)
Nikon D2H (45ms/46ms)
Nikon D100 (100ms/100ms)
Nikon D2X (44ms/44ms)
Canon 10D (104ms/146ms)
Canon 20D (77ms/110ms)
Canon 5D (78ms/133ms)
Canon 1D (39ms/54ms)
Canon 1D Mk II (54ms/75ms)
Canon 1DS (59ms/83ms)
Canon 300D (142ms/248ms)
Canon 350D (95ms/168ms)
Fujifilm S2 Pro (162ms/220ms)
Olympus E1 (76.5ms/173ms)
Olympus E-300 (100ms/370ms)
Olympus E-500 (95ms/320ms)
Pentax *ist-D (130ms/180ms)
Sigma SD10 (111ms/no timing for manual focus)
Kodak DCS Pro 14n (119ms/128ms)
Minolta D7D (117ms/253ms)

camera (autofocus lag) lens
Nikon D50 (265ms) AF-S 18-55
Nikon D70 (342/486ms) AF-S 18-70
Nikon D100 (150ms) 24-85
Nikon D2X (44ms) ?
Canon 300D (250/278ms) EF-S 18-55
Canon 20D (160/150ms) EF-S 18-55 USM
Canon 5D (149ms) 100mm macro
Canon 350D (240ms/220ms) EF-S 18-55
Canon 1D Mk II (200ms/240ms) 16-35 F2.8L
Canon 1DS (93ms) 100mm macro
Olympus E-300 (360/370ms) Kit 14-45
Olympus E-500 (370ms/380ms) ?
Minolta D7D (266ms) ?

 

[OlyFlyer]

maybe it's a purposely built-in delay to ensure that everything is ready before the shutter opens? even if you fixed all the parameters, the body still needs to inform the lens what aperture is to be used and to close down if necessary (please correct me if I'm wrong).

True, true. Still, 250ms i a very long time for that. Even older film cameras has to do that, and in my old OM2 the delay was not that long. That camera is actually electronic in that matter, there is a solenoid (electric magnet) that handles that part. These solenoids supposed to be slower than todays electronic.

Another website does its own testing of such lags - have to say the E500 isn't very fast among the current DSLRs:
http://www.imaging-resource.com/PRODS/EV500/EV500A9.HTM

Interesting reading. Some of his figures are similar to mine, some conclusions he make are however wrong. Maybe he just did not make enough measurments, maybe he made a mistake. Examples are the delays Cycle Time, max/min resolution. In reality, that largly depends on the CF used. Delay is practically 0 if Sandisk Extreme III CF is used. Buffer is flashed without you noticing it. I think that is valid for all JPEG resolutions. He says that '...but the shutter response in manual focus mode is surprisingly slow, with a delay of 0.32 second (almost the same as for full-AF mode).' That is also about the same conclusion as I did. The time difference beween his and my measurments may depend on camera individuality or measurment errors/methods. I think actually, my measurments regarding manual settings are more accurate than his. Also, I think I fired more shots than he did and therfore got more accurate avarages.

 

[mpenza]

He used a kingston 50x (more than 3 times slower than Extreme III!), so maybe that explains the slower write speeds ;p guess he has to, in order for his earlier reviews to be more easily comparable and maybe also to have measureable numbers (might not too fast to get a timing ;p).

I take his numbers more for relative comparison across the cameras he has reviewed rather than absolute figures (since the same methodology/equipment was used). There could be some calibration error that cause some deviations from yours and timings done by others (e.g. his numbers are different vs www.dpreview.com).

 

[OlyFlyer]

In fact, Oly recomends Extreme III if you really want to take advantage of the Oly processor speed. Oly Sweden gave away Sandisk Ultra II 512MB CF with every E-500 when I bought mine. Even that is faster than his Kingston. I guarantee, you notice the difference, even between my two CFs. So, I think any serious tester should eliminate the CF speed as far as that is possible. Otherways the figures are all wrong.

 

[mpenza]

As I mentioned, he may want to have his results comparable to his earlier reviews or may not have the equipment to measure very fast write speeds. Dun think u can conclude he's not serious given the extent he goes to to make test results comparable (check out the sample test images which remained the same for several years!) ;p

I use different sources for memory card I/O performance (www.imaging-resource.com is still alright for AF speeds though). Usually, I refer to Rob Galbraith's page but he doesn't cover Olympus cameras:
http://www.robgalbraith.com/bins/multi_page.asp?cid=6007

DPReview has some details on E-500 card performance.... http://www.dpreview.com/reviews/olympuse500/page10.asp

E-500 writes RAW files really fast on the Extreme III (9.85MB/s - faster than top pro cameras from Nikon and Olympus)!!!! JPEG write speed is quite fast too (4.68MB/s). Impressive!

 

[OlyFlyer]

Sorry if you misunderstood me, I don't mean that the guy is not serious, just that to make a fair comparision you must let time and development advance and use what's available. When he started his work there wasn't any Extreme III CFs, so he used other, slower cards. Also cameras were slower. There is also an Oly recomendation to use Extreme III cards if you want speed. So to make a fair test, according to me, you must use the best CF. I think these guys ARE doing great and serious jobs, but things can be measured differently with results that are not always fair.

Regarding DPReview, I don't understand why they don't have faster, better results for Sandisk CFs. Extreme III has write and read speed of 20MB/s, while Ultra II has 10MB/s, so RAW continuous burst writes should be faster. Again, I don't question the result, just think it is strange. If measured correctly, than it must mean that the camera somehow does not like Ultra II CFs, since the camera must have at least 9MB/s flush speed with Extreme III, it should have the same result with Ultra II. Or, Sandisk is cheating when they declare speeds. I did not measure flush speed, but I can assure you that my feeling say also Ultra II is half as fast as Extreme III.

 

[mpenza]

What they have are observations. Perhaps, it's important to understand the conditions and assumptions for the specified write/read speeds (I dun really know too) and cards specified to be the same speeds by different manufacturers perform differently in practical use (and some cards work faster on some cameras but slower on others).