No, that was a separate test, and I was using the LCD as an eye-level EVF. And, yes, I was unable to use it effectively because of EVF lag, as I clearly demonstrated.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
D
Donald B
Guest
read the next postDo you have dementia or something?your kidding arnt you ;-) i can write a celestial navigation program on excel with no references needed , this is a simple task i have asked.And how am I going to make a spreadsheet from that useless information?shot this afternoon a 2sec sequence of a car traveling at 80klm per hour with a 120mm lens (ff eq) car was 40 meters away shot 20 images with the car filling half the frame and perfectly in the center, how many rpm was i panning at ? and how fast would the car have to be traveling to exit the vf ? and what would be the rpm of my body panningOkay, shoot video this time, use a millisecond timer, and give me the video and I'll process it.show us the math on an excell spread sheet for all us to see and ad some of our own numbers in
The question is "how much lag does your EVF have", and I tried to estimate that from the lousy pictures you provided. You asked for the math so I asked for you to shoot a video of a millisecond timer so I could calculate it. Now you're talking about a car. What the heck are you talking about?
Don
Its on another topic, and you're still mixing in human reaction time, which is irrelevant. According to that, this video I took couldn't have been taken. That plane was 40 feet from me at closest approach going about 200mph.read the next postDo you have dementia or something?your kidding arnt you ;-) i can write a celestial navigation program on excel with no references needed , this is a simple task i have asked.And how am I going to make a spreadsheet from that useless information?shot this afternoon a 2sec sequence of a car traveling at 80klm per hour with a 120mm lens (ff eq) car was 40 meters away shot 20 images with the car filling half the frame and perfectly in the center, how many rpm was i panning at ? and how fast would the car have to be traveling to exit the vf ? and what would be the rpm of my body panningOkay, shoot video this time, use a millisecond timer, and give me the video and I'll process it.show us the math on an excell spread sheet for all us to see and ad some of our own numbers in
The question is "how much lag does your EVF have", and I tried to estimate that from the lousy pictures you provided. You asked for the math so I asked for you to shoot a video of a millisecond timer so I could calculate it. Now you're talking about a car. What the heck are you talking about?
Oh...I got plenty of good stills of that same plane going the same speed, pretty tightly framed, using an OVF:
--
Lee Jay
It all matters. You should learn to listen.
BarelyAmateur
Leading Member
- Messages
- 505
- Reaction score
- 454
Very close, but this isn't quite the right way. The best thing would be to take a high speed camera that is capable of around 240fps. This will give you a measurement to the nearest ~4ms, which should be sufficient.
Then point your camera at a screen where something is changing, just as you did, and take a video. Then look at the video frame by frame and measure how long it took for a change in the source to be reflected as a change in the target.
I can tell you right now that this is almost certainly not going to be faster than about four frames (16ms). That's one frame at 60fps. I doubt it will be slower than 50ms. Either way, it's not enough to materially effect panning except for objects that are extremely close to the camera moving extremely fast. (100 km/h * 50ms = ~1 meter).
Then point your camera at a screen where something is changing, just as you did, and take a video. Then look at the video frame by frame and measure how long it took for a change in the source to be reflected as a change in the target.
I can tell you right now that this is almost certainly not going to be faster than about four frames (16ms). That's one frame at 60fps. I doubt it will be slower than 50ms. Either way, it's not enough to materially effect panning except for objects that are extremely close to the camera moving extremely fast. (100 km/h * 50ms = ~1 meter).
D
Donald B
Guest
The problem is Lea is your shots are composition aly not good as they should be flying into the frame not out of it. So the small delay would have enhanced your images
Don
Don
BarelyAmateur
Leading Member
- Messages
- 505
- Reaction score
- 454
High res clock would be pointless in this context because the screen will only display updates once per 16ms. (Unless it's a 144hz screen or something.)
D
Donald B
Guest
thank you very much for the reply, i knew someone with math skill would chim in soon.
Don
D
Donald B
Guest
Astrotripper
Veteran Member
- Messages
- 8,676
- Solutions
- 14
- Reaction score
- 12,107
You're not going to get an EVF lag by measuring lag of the back screen.
I have no proof for the statement that follows, but I'd bet my E-M1 II that the lag on the back screen is a lot higher than that of EVF.
Heck, the input lag on that back screen LCD panel by itself is probably higher than the lag of the whole data pipeline for the EVF (ie. from sensor readout to all the way to the displayed image).
I also don't know how you could reliably measure this stuff by yourself without some elaborate setup. Even just taking a photo like you did here means you need to account for a rolling shutter effect (and yes, even with mechanical one).
I have no proof for the statement that follows, but I'd bet my E-M1 II that the lag on the back screen is a lot higher than that of EVF.
Heck, the input lag on that back screen LCD panel by itself is probably higher than the lag of the whole data pipeline for the EVF (ie. from sensor readout to all the way to the displayed image).
I also don't know how you could reliably measure this stuff by yourself without some elaborate setup. Even just taking a photo like you did here means you need to account for a rolling shutter effect (and yes, even with mechanical one).
Astrotripper
Veteran Member
- Messages
- 8,676
- Solutions
- 14
- Reaction score
- 12,107
Actually, maybe it would not be as complex as I initially thought. Super slo-mo modes on moderns smartphones should make the problem of rolling shutter minimal (and it could be eliminated completely by setting up the scene in a specific way). And you could take a video of an RGB LED alternating through different states. It would be quite simple to create a setup where you could capture both the EVF and the LED itself. And you could control the frequency of LED changes to get the best possible measurement.
Something like that.
Martin.au
Forum Pro
If you are doing timing tests, then the fact that the bottom of a photo is taken a variable amount of time after the top of the photo absolutely matters (depending on camera and settings).
New Day Rising
Veteran Member
AlwynS
Senior Member
Seriously??!! This whole storm in a teacup is based on the fact that he is not familiar with the generally accepted photographic colloquial distinction between an EVF and an LCD? Could you point me at a post where this is made public please?
Over and above anything else: attempting to track anything while holding a camera at arm's length to see the rear screen is a complete and utter waste of time. This is a far, far greater issue than simple display lag.
Holy cow, talk about absurd....
Since when should all those happy with EVFs need to convince DSLR shooters to move? I am happy if they stay with the DSLRs and they have every right to prefer it.
New Day Rising
Veteran Member
mamallama
Forum Pro
It is included in the shutter lag measurements in the performance tests results posed in the DSLR camera reviews in the Imaging Resources web site.
Last edited:
DavidArmenPhoto
Forum Enthusiast
- Messages
- 370
- Solutions
- 1
- Reaction score
- 179
Literally the best post. Mirrorless EVF will ALWAYS have a lag, because of the laws of physics. We may decrease that lag to extremely small amounts in the years to come but it will never ever be zero. Just want to clarify that.
mamallama
Forum Pro
Neither will an OVF lag ever be zero, literally. :-D
They both may become inconsequential in the future, however.