Personal View site logo
Make sure to join PV Telegram channel! Perfect to keep up with community on your smartphone.
ETC Mode Discussion
  • I own only a GH1 (no etc mode) but I'm planning to buy a GF2, which has only x3.2 ETC mode (while GH2 has more options on ETC). I like very much old fast lenses like c-mount tamron 4-12 zooms, pentax 6-41, it would be great to use them (they're very cheap, you know) on GF2, however most of the demos I've found on the net are GH2-specific,and I don't know what are ETC drawbacks. A x3.2 could be too "extreme" in terms of video quality? It would be a good idea to use a c-mount 4-12 tamron f 1.0 on a GF2 x3.2 ETc (24-72mm!!) or video quality would be worse than a common kit lens panasonic 14-42 f3.5?

  • 105 Replies sorted by
  • @AlbertZ There aren't actually a lot of ETC mode options on the GH2. The only option is whether ETC crops to 1280x720 or 1920x1080, so it is pretty much determined by the output resolution you want.

  • ETC mode adds grain (noise) at lower ISO, but the advantage is you can be at 160 all the time with fast lenses found in c mount.. one other benefit/disadvantage is it acts like a small sensor camera.. less shallow DOF than full frame, but also has a small sensor advantage , less jello (virtual none) with fast camera motion.

    SO it has benefits and drawbacks, I can say I love ETC and use it in production.. for those special shots that I need either really fast (f1.0) lenses or super wide fast lenses, like the Tamron 4-12mm (I love that lens)

    ETC mode makes nice video if you can keep the ISO very low.. Typically I'm at 160 ISO all the time in ETC mode. shooting like this is can intermix with full framed images with out a problem.. i can tell the difference but clients can't

    i'm speaking about use of etc on gh2, but gfx would be similar.. have not really done allot of etc on the gf bodies

  • @Ebacherville Thanks for the balanced summing-up of ETC!

  • GFx supports 720p ETC only.

  • Anyone else got a dead pixel in ETC? That's a pretty big drawback!

    Also, you can't shoot ETC mode if you're running an external monitor. That's really annoying, but not so much a drawback to ETC in itself.

  • no dead pixels that ive seen in ETC, but banding seems to be worse in ETC .. but only @ higher ISO, but typically when I'm in ETC I'm at the lowest ISO possible with very fast lenses. My main use for ETC is ultra wide shots with the 4-12mm Tamron, or a rare need for a para focal zoom in low light using one of my fast zooms.

  • @Ebacherville

    "one other benefit/disadvantage is it acts like a small sensor camera.. less shallow DOF than full frame, but also has a small sensor advantage , less jello (virtual none) with fast camera motion."

    I don't believe this is correct. Jello is affected by sensor readout speed. I don't think the sensor reads any faster with the cropped frame.

    In reality Jello is not made worse or better with ETC. However, the ETC's zoomed in frame can make jello more apparent especially in slow motion.

  • Thanks for your reply! I'd like to use c-mount lenses and low iso, then.@Ebacherville Of course, there's no need to use high ISO when we have cheap and super-fast c-mount lenses ;) I've found great videos on the net, the only issue is that is very difficul to find GF2/3 + ETC demos..

  • @mpgxsvcd

    Intuitively jello & rolling shutter effects wouldn’t be improved on the sensor cropped image compared to its non-cropped image’s “cropped” center part that matches to the ETC’s framing.

    But ETC would show less jello/rolling shutter effects than using non-ETC mode at the same framing. E.g. 20mm with ETC vs 52mm lens with no ETC.

  • Jello is affected by sensor readout speed. I don't think the sensor reads any faster with the cropped frame.

    It is complicated topic. As we don't really know how reading is made on low level.
    It is clear that less pixels must be read in cropped mode. So, with same speed readout will be faster for each line.

  • It's accepted knowledge that ETC yields less jello. If anyone doubts it, test it. I experimented quite a bit for a chase sequence and the difference is dramatic. Although, I don't believe it completely eliminates it. I dunno if it's perceived or actual, all I know is in ETC you get a lot less jello. It's a major feature of the GH2 in my mind.

  • @brianluce

    As I pointed adove, while read speed is the same, frame reading must be faster for crop mode. Saving can be due to less horizontal lines required to be read (as you have exact one line for one line of image in crop mode).

  • @Vitaliy I understand, I just wanted to highlight that capability because it seems no one mentions it much. For hand held action/crash sequence, it can make a huge difference. I never used it much for tele, if I want tele I change the lens.

  • @brianluce Hand held and ETC? Good news! ;) It could be interesting a video comparision: ETC-jello, no ETC-jello ;)

  • sky banding issue is the worst scenario with ETC, it is like flickering light, you must find right shutter speed, and even then it is very noticeable, so most of the shoots with sky in it, done with ETC are junk

  • @Vitaliy_Kiselev

    You are right. The issue is that we don't know what the sensor does. I guess only testing can settle this. Although this is hard to accurately test.

    I guess we have to test the 14-140mm lens at 35mm with the 720p @ 60 FPS 3.9x ETC and then test the same thing at 720p @ 60 FPS for a true 140mm lens.

    I am thinking of testing this by filming cars on the interstate. There is an overpass I can film from and track them from about 1 mile away to within 100 feet directly below me. If I film passing trucks their speed should be relatively constant because the cops are all over that section of road.

    You can determine the level of jello with a passing truck by the angle of the back of the truck. The more jello the less perpendicular the back will be to the top of the truck.

  • @oneday really?? why is that?

  • Shoot an electric fan at constant speed, the amount of deterioration of the blades is a very good indicator for RS.

  • @AlbertZ what's the Pentax 6-41?

  • @nomad

    Shooting the trucks worked very well. I will post the videos when I get home.

  • This video shows the Jello Effect when ETC mode is used. I used the 14-140mm @ 35mm plus the 3.9x 720p ETC and the 14-140mm @ 140mm to give the same equivalent field of view.

    I filmed passing trucks at the same speed with 1/250th shutter speed. I could not tell any difference in the angle of distortion between the two clips. A lot of people think that rolling shutters only affect the video if the camera is moving. It will actually slant vertical lines if the subject is moving fast enough horizontally even if the camera is still.

    I do not believe that using ETC mode over just the lens has any affect on the rolling shutter jello effect. Theoretically sensor readout speed is the only thing that could affect it. Using ETC mode should reduce the number of pixels that the camera reads. However, that does not appear to help at all. Instead it appears to read the entire sensor in the same time that it takes to read it without using ETC. Maybe it is just throwing out the data that the ETC mode doesn't need?

    Sorry it is late and I didn't know where the original thread was so I had originally posted this as its own topic.

  • @mpgxsvcd

    Great test. This myth can be now put to bed.

    @subco

    rolling shutter will be present regardless of using ETC or not

    Exactly! Also... Changing the shutter speed values will only make the Jello crisp or blurry.

    can be otherwise expressed as, "Changing the shutter speed values will only make the movement smooth or stuttering".

  • Good test indeed, @mpgxsvcd

  • Am I the only one seeing slightly less rolling shutter effect from ETC? But the improvement seems so little that it doesn't matter. Thanks for the test, @mpgxsvcd.

  • @stonebat

    That is correct. If there is a difference then it is negligible. However, the more important thing is “why isn’t there a bigger difference?”. If the sensor is read at the same rate then it stands to reason that it should read the entire sensor faster if it is only reading a portion of the sensor.

    I don’t fully understand how these sensors are read. I will leave that up to you experts. However, I do believe that it should be possible to virtually eliminate rolling shutter by using a fast readout for the sensor(1/250th like in the Nikon 1 series cameras) and using a cropped sensor mode.